2024 in Vision [LS Live @ NeurIPS]

Happy holidays! We’ll be sharing snippets from Latent Space LIVE! https://lu.ma/LSLIVE through the break bringing you the best of 2024! We want to express our deepest appreciation to event sponsors AWS https://www.linkedin.com/in/aaronamelgar/, Daylight Computer https://daylightcomputer.com/, Thoth.ai https://x.com/_thothAI/status/1866894234953060666, StrongCompute https://x.com/strongcompute?lang=bn, Notable Capital https://www.linkedin.com/in/laura-hamilton/, and most of all all our LS supporters https://www.latent.space/subscribe?utm_source=menu&simple=true&next=https%3A%2F%2Fwww.latent.space%2F who helped fund the gorgeous venue and A/V production! For NeurIPS last year https://www.latent.space/p/neurips-2023-papers we did our standard conference podcast coverage interviewing selected papers (that we have now also done for ICLR https://www.latent.space/p/iclr-2024-benchmarks-agents?utm_source=publication-search and ICML https://www.latent.space/p/icml-2024-video-robots), however we felt that we could be doing more to help AI Engineers 1) get more industry-relevant content, and 2) recap 2024 year in review from experts. As a result, we organized the first Latent Space LIVE!, our first in person miniconference, at NeurIPS 2024 in Vancouver. The single most requested domain was computer vision, and we could think of no one better to help us recap 2024 than our friends at Roboflow, who was one of our earliest guests in 2023 https://www.latent.space/p/segment-anything-roboflow?utm_source=publication-search and had one of this year’s top episodes in 2024 https://www.latent.space/p/sam2?utm_source=publication-search again. Roboflow has since raised a $40m Series B https://blog.roboflow.com/series-b/#2! Links Their slides are here https://docs.google.com/presentation/d/1iRUBWw_5srDQo1NBl4REJTD_BHE1jmphF3-nzLIsPg4/edit?usp=drivesdk: All the trends and papers they picked: * Isaac Robinson https://www.linkedin.com/in/robinsonish/ * Sora https://openai.com/index/sora/ (see our Video Diffusion pod https://www.latent.space/p/icml-2024-video-robots) - extending diffusion from images to video * SAM 2: Segment Anything in Images and Videos https://arxiv.org/abs/2408.00714 (see our SAM2 pod https://latent.space/p/sam2) - extending prompted masks to full video object segmentation * DETR Dominancy: DETRs show Pareto improvement over YOLOs * RT-DETR: DETRs Beat YOLOs on Real-time Object Detection https://arxiv.org/abs/2304.08069 * LW-DETR:A Transformer Replacement to YOLO for Real-Time Detection https://arxiv.org/abs/2406.03459 * D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement https://arxiv.org/abs/2410.13842 * Peter Robicheaux https://www.linkedin.com/in/peter-robicheaux-01958813b/ * MMVP (Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs) https://arxiv.org/abs/2401.06209 * * Florence 2 (Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks) https://arxiv.org/abs/2311.06242 * PalíGemma / PaliGemma 2 * PaliGemma: A versatile 3B VLM for transfer https://arxiv.org/abs/2407.07726 * PaliGemma 2: A Family of Versatile VLMs for Transfer https://arxiv.org/abs/2412.03555 * AlMv2 (Multimodal Autoregressive Pre-training of Large Vision Encoders) https://arxiv.org/abs/2411.14402 * Vik Korrapati - Moondream https://github.com/vikhyat/moondream Full Talk on YouTube Want more content like this? Like and subscribe to stay updated on our latest talks, interviews, and podcasts. Transcript/Timestamps [00:00:00] Intro [00:00:05] AI Charlie: welcome to Latent Space Live, our first mini conference held at NeurIPS 2024 in Vancouver. This is Charlie, your AI co host. When we were thinking of ways to add value to our academic conference coverage, we realized that there was a lack of good talks, just recapping the best of 2024, going domain by domain. [00:00:36] AI Charlie: We sent out a survey to the over 900 of you. who told us what you wanted, and then invited the best speakers in the Latent Space Network to cover each field. 200 of you joined us in person throughout the day, with over 2, 200 watching live online. Our second featured keynote is The Best of Vision 2024, with Peter Robichaud and Isaac [00:01:00] Robinson of Roboflow, with a special appearance from Vic Corrapati of Moondream. [00:01:05] AI Charlie: When we did a poll of our attendees, the highest interest domain of the year was vision. And so our first port of call was our friends at Roboflow. Joseph Nelson helped us kickstart our vision coverage in episode 7 last year, and this year came back as a guest host with Nikki Ravey of Meta to cover segment Anything 2. [00:01:25] AI Charlie: Roboflow have consistently been the leaders in open source vision models and tooling. With their SuperVision library recently eclipsing PyTorch's Vision library. And Roboflow Universe hosting hundreds of thousands of open source vision datasets and models. They have since announced a 40 million Series B led by Google Ventures. [00:01:46] AI Charlie: Woohoo. [00:01:48] Isaac's picks [00:01:48] Isaac Robinson: Hi, we're Isaac and Peter from Roboflow, and we're going to talk about the best papers of 2024 in computer vision. So, for us, we defined best as what made [00:02:00] the biggest shifts in the space. And to determine that, we looked at what are some major trends that happened and what papers most contributed to those trends. [00:02:09] Isaac Robinson: So I'm going to talk about a couple trends, Peter's going to talk about a trend, And then we're going to hand it off to Moondream. So, the trends that I'm interested in talking about are These are a major transition from models that run on per image basis to models that run using the same basic ideas on video. [00:02:28] Isaac Robinson: And then also how debtors are starting to take over the real time object detection scene from the YOLOs, which have been dominant for years. [00:02:37] Sora, OpenSora and Video Vision vs Generation [00:02:37] Isaac Robinson: So as a highlight we're going to talk about Sora, which from my perspective is the biggest paper of 2024, even though it came out in February. Is the what? [00:02:48] Isaac Robinson: Yeah. Yeah. So just it's a, SORA is just a a post. So I'm going to fill it in with details from replication efforts, including open SORA and related work, such as a stable [00:03:00] diffusion video. And then we're also going to talk about SAM2, which applies the SAM strategy to video. And then how debtors, These are the improvements in 2024 to debtors that are making them a Pareto improvement to YOLO based models. [00:03:15] Isaac Robinson: So to start this off, we're going to talk about the state of the art of video generation at the end of 2023, MagVIT MagVIT is a discrete token, video tokenizer akin to VQ, GAN, but applied to video sequences. And it actually outperforms state of the art handcrafted video compression frameworks. [00:03:38] Isaac Robinson: In terms of the bit rate versus human preference for quality and videos generated by autoregressing on these discrete tokens generate some pretty nice stuff, but up to like five seconds length and, you know, not super detailed. And then suddenly a few months later we have this, which when I saw it, it was totally mind blowing to me. [00:03:59] Isaac Robinson: 1080p, [00:04:00] a whole minute long. We've got light reflecting in puddles. That's reflective. Reminds me of those RTX demonstrations for next generation video games, such as Cyberpunk, but with better graphics. You can see some issues in the background if you look closely, but they're kind of, as with a lot of these models, the issues tend to be things that people aren't going to pay attention to unless they're looking for. [00:04:24] Isaac Robinson: In the same way that like six fingers on a hand. You're not going to notice is a giveaway unless you're looking for it. So yeah, as we said, SORA does not have a paper. So we're going to be filling it in with context from the rest of the computer vision scene attempting to replicate these efforts. So the first step, you have an LLM caption, a huge amount of videos. [00:04:48] Isaac Robinson: This, this is a trick that they introduced in Dolly 3, where they train a image captioning model to just generate very high quality captions for a huge corpus and then train a diffusion model [00:05:00] on that. Their Sora and their application efforts also show a bunch of other steps that are necessary for good video generation. [00:05:09] Isaac Robinson: Including filtering by aesthetic score and filtering by making sure the videos have enough motion. So they're not just like kind of the generators not learning to just generate static frames. So. Then we encode our video into a series of space time latents. Once again, SORA, very sparse in details. [00:05:29] Isaac Robinson: So the replication related works, OpenSORA actually uses a MAG VIT V2 itself to do this, but swapping out the discretization step with a classic VAE autoencoder framework. They show that there's a lot of benefit from getting the temporal compression, which makes a lot of sense as the Each sequential frames and videos have mostly redundant information. [00:05:53] Isaac Robinson: So by compressing against, compressing in the temporal space, you allow the latent to hold [00:06:00] a lot more semantic information while avoiding that duplicate. So, we've got our spacetime latents. Possibly via, there's some 3D VAE, presumably a MAG VATV2 and then you throw it into a diffusion transformer. [00:06:19] Isaac Robinson: So I think it's personally interesting to note that OpenSORA is using a MAG VATV2, which originally used an autoregressive transformer decoder to model the latent space, but is now using a diffusion diffusion transformer. So it's still a transformer happening. Just the question is like, is it? [00:06:37] Isaac Robinson: Parameterizing the stochastic differential equation is, or parameterizing a conditional distribution via autoregression. It's also it's also worth noting that most diffusion models today, the, the very high performance ones are switching away from the classic, like DDPM denoising diffusion probability modeling framework to rectified flows. [00:06:57] Isaac Robinson: Rectified flows have a very interesting property that as [00:07:00] they converge, they actually get closer to being able to be sampled with a single step. Which means that in practice, you can actually generate high quality samples much faster. Major problem of DDPM and related models for the past four years is just that they require many, many steps to generate high quality samples. [00:07:22] Isaac Robinson: So, and naturally, the third step is throwing lots of compute at the problem. So I didn't, I never figured out how to manage to get this video to loop, but we see very little compute, medium compute, lots of compute. This is so interesting because the the original diffusion transformer paper from Facebook actually showed that, in fact, the specific hyperparameters of the transformer didn't really matter that much. [00:07:48] Isaac Robinson: What mattered was that you were just increasing the amount of compute that the model had. So, I love how in the, once again, little blog posts, they don't even talk about [00:08:00] like the specific hyperparameters. They say, we're using a diffusion transformer, and we're just throwing more compute at it, and this is what happens. [00:08:08] Isaac Robinson: OpenSora shows similar results. The primary issue I think here is that no one else has 32x compute budget. So we end up with these we end up in the middle of the domain and most of the related work, which is still super, super cool. It's just a little disappointing considering the context. So I think this is a beautiful extension of the framework that was introduced in 22 and 23 for these very high quality per image generation and then extending that to videos. [00:08:39] Isaac Robinson: It's awesome. And it's GA as of Monday, except no one can seem to get access to it because they keep shutting down the login. [00:08:46] SAM and SAM2 [00:08:46] Isaac Robinson: The next, so next paper I wanted to talk about is SAM. So we at Roboflow allow users to label data and train models on that data. Sam, for us, has saved our users 75 years of [00:09:00] labeling time. [00:09:00] Isaac Robinson: We are the, to the best of my knowledge, the largest SAM API that exists. We also, SAM also allows us to have our users train just pure bounding box regression models and use those to generate high quality masks which has the great side effect of requiring less training data to have a meaningful convergence. [00:09:20] Isaac Robinson: So most people are data limited in the real world. So anything that requires less data to get to a useful thing is that super useful. Most of our users actually run their object per frame object detectors on every frame in a video, or maybe not most, but many, many. And so Sam follows into this category of taking, Sam 2 falls into this category of taking something that really really works and applying it to a video which has the wonderful benefit of being plug and play with most of our Many of our users use cases. [00:09:53] Isaac Robinson: We're, we're still building out a sufficiently mature pipeline to take advantage of that, but it's, it's in the works. [00:10:00] So here we've got a great example. We can click on cells and then follow them. You even notice the cell goes away and comes back and we can still keep track of it which is very challenging for existing object trackers. [00:10:14] Isaac Robinson: High level overview of how SAM2 works. We there's a simple pipeline here where we can give, provide some type of prompt and it fills out the rest of the likely masks for that object throughout the rest of the video. So here we're giving a bounding box in the first frame, a set of positive negative points, or even just a simple mask. [00:10:36] Isaac Robinson: I'm going to assume people are somewhat familiar with SAM. So I'm going to just give a high level overview of how SAM works. You have an image encoder that runs on every frame. SAM two can be used on a single image, in which case the only difference between SAM two and SAM is that image encoder, which Sam used a standard VIT [00:11:00] Sam two replaced that with a hara hierarchical encoder, which gets approximately the same results, but leads to a six times faster inference, which is. [00:11:11] Isaac Robinson: Excellent, especially considering how in a trend of 23 was replacing the VAT with more efficient backbones. In the case where you're doing video segmentation, the difference is that you actually create a memory bank and you cross attend the features from the image encoder based on the memory bank. [00:11:31] Isaac Robinson: So the feature set that is created is essentially well, I'll go more into it in a couple of slides, but we take the features from the past couple frames, plus a set of object pointers and the set of prompts and use that to generate our new masks. Then we then fuse the new masks for this frame with the. [00:11:57] Isaac Robinson: Image features and add that to the memory bank. [00:12:00] It's, well, I'll say more in a minute. The just like SAM, the SAM2 actually uses a data engine to create its data set in that people are, they assembled a huge amount of reference data, used people to label some of it and train the model used the model to label more of it and asked people to refine the predictions of the model. [00:12:20] Isaac Robinson: And then ultimately the data set is just created from the engine Final output of the model on the reference data. It's very interesting. This paradigm is so interesting to me because it unifies a model in a dataset in a way that is very unique. It seems unlikely that another model could come in and have such a tight. [00:12:37] Isaac Robinson: So brief overview of how the memory bank works, the paper did not have a great visual, so I'm just, I'm going to fill in a bit more. So we take the last couple of frames from our video. And we take the last couple of frames from our video attend that, along with the set of prompts that we provided, they could come from the future, [00:13:00] they could come from anywhere in the video, as well as reference object pointers, saying, by the way, here's what we've found so far attending to the last few frames has the interesting benefit of allowing it to model complex object motion without actually [00:13:18] Isaac Robinson: By limiting the amount of frames that you attend to, you manage to keep the model running in real time. This is such an interesting topic for me because one would assume that attending to all of the frames is super essential, or having some type of summarization of all the frames is super essential for high performance. [00:13:35] Isaac Robinson: But we see in their later ablation that that actually is not the case. So here, just to make sure that there is some benchmarking happening, we just compared to some of the stuff that's came out prior, and indeed the SAM2 strategy does improve on the state of the art. This ablation deep in their dependencies was super interesting to me. [00:13:59] Isaac Robinson: [00:14:00] We see in section C, the number of memories. One would assume that increasing the count of memories would meaningfully increase performance. And we see that it has some impact, but not the type that you'd expect. And that it meaningfully decreases speed, which justifies, in my mind, just having this FIFO queue of memories. [00:14:20] Isaac Robinson: Although in the future, I'm super interested to see A more dedicated summarization of all of the last video, not just a stacking of the last frames. So that another extension of beautiful per frame work into the video domain. [00:14:42] Realtime detection: DETRs > YOLO [00:14:42] Isaac Robinson: The next trend I'm interested in talking about is this interesting at RoboFlow, we're super interested in training real time object detectors. [00:14:50] Isaac Robinson: Those are bread and butter. And so we're doing a lot to keep track of what is actually happening in that space. We are finally starting to see something change. So, [00:15:00] for years, YOLOs have been the dominant way of doing real time object detection, and we can see here that they've essentially stagnated. [00:15:08] Isaac Robinson: The performance between 10 and 11 is not meaningfully different, at least, you know, in this type of high level chart. And even from the last couple series, there's not. A major change so YOLOs have hit a plateau, debtors have not. So we can look here and see the YOLO series has this plateau. And then these RT debtor, LW debtor, and Define have meaningfully changed that plateau so that in fact, the best Define models are plus 4. [00:15:43] Isaac Robinson: 6 AP on Cocoa at the same latency. So three major steps to accomplish this. The first RT deditor, which is technically a 2023 paper preprint, but published officially in 24, so I'm going to include that. I hope that's okay. [00:16:00] That is showed that RT deditor showed that we could actually match or out speed YOLOs. [00:16:04] Isaac Robinson: And then LWdebtor showed that pre training is hugely effective on debtors and much less so on YOLOs. And then DeFine added the types of bells and whistles that we expect from these types, this, this arena. So the major improvements that RTdebtor shows was Taking the multi scale features that debtors typically pass into their encoder and decoupling them into a much more efficient transformer encoder. [00:16:30] Isaac Robinson: The transformer is of course, quadratic complexity. So decreasing the amount of stuff that you pass in at once is super helpful for increasing your runtime or increasing your throughput. So that change basically brought us up to yellow speed and then they do a hardcore analysis on. Benchmarking YOLOs, including the NMS step. [00:16:54] Isaac Robinson: Once you once you include the NMS in the latency calculation, you see that in fact, these debtors [00:17:00] are outperforming, at least this time, the the, the YOLOs that existed. Then LW debtor goes in and suggests that in fact, the frame, the huge boost here is from pre training. So, this is the define line, and this is the define line without pre training. [00:17:19] Isaac Robinson: It's within range, it's still an improvement over the YOLOs, but Really huge boost comes from the benefit of pre training. When YOLOx came out in 2021, they showed that they got much better results by having a much, much longer training time, but they found that when they did that, they actually did not benefit from pre training. [00:17:40] Isaac Robinson: So, you see in this graph from LWdebtor, in fact, YOLOs do have a real benefit from pre training, but it goes away as we increase the training time. Then, the debtors converge much faster. LWdebtor trains for only 50 epochs, RTdebtor is 60 epochs. So, one could assume that, in fact, [00:18:00] the entire extra gain from pre training is that you're not destroying your original weights. [00:18:06] Isaac Robinson: By relying on this long training cycle. And then LWdebtor also shows superior performance to our favorite data set, Roboflow 100 which means that they do better on the real world, not just on Cocoa. Then Define throws all the bells and whistles at it. Yellow models tend to have a lot of very specific complicated loss functions. [00:18:26] Isaac Robinson: This Define brings that into the debtor world and shows consistent improvement on a variety of debtor based frameworks. So bring these all together and we see that suddenly we have almost 60 AP on Cocoa while running in like 10 milliseconds. Huge, huge stuff. So we're spending a lot of time trying to build models that work better with less data and debtors are clearly becoming a promising step in that direction. [00:18:56] Isaac Robinson: The, what we're interested in seeing [00:19:00] from the debtors in this, this trend to next is. Codetter and the models that are currently sitting on the top of the leaderboard for large scale inference scale really well as you switch out the backbone. We're very interested in seeing and having people publish a paper, potentially us, on what happens if you take these real time ones and then throw a Swingy at it. [00:19:23] Isaac Robinson: Like, do we have a Pareto curve that extends from the real time domain all the way up to the super, super slow but high performance domain? We also want to see people benchmarking in RF100 more, because that type of data is what's relevant for most users. And we want to see more pre training, because pre training works now. [00:19:43] Isaac Robinson: It's super cool. [00:19:48] Peter's Picks [00:19:48] Peter Robicheaux: Alright, so, yeah, so in that theme one of the big things that we're focusing on is how do we get more out of our pre trained models. And one of the lenses to look at this is through sort of [00:20:00] this, this new requirement for like, how Fine grained visual details and your representations that are extracted from your foundation model. [00:20:08] Peter Robicheaux: So it's sort of a hook for this Oh, yeah, this is just a list of all the the papers that I'm going to mention I just want to make sure I set an actual paper so you can find it later [00:20:18] MMVP (Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs) [00:20:18] Peter Robicheaux: Yeah, so sort of the big hook here is that I make the claim that LLMs can't see if you go to if you go to Claude or ChatGPT you ask it to see this Watch and tell me what time it is, it fails, right? [00:20:34] Peter Robicheaux: And so you could say, like, maybe, maybe the Like, this is, like, a very classic test of an LLM, but you could say, Okay, maybe this, this image is, like, too zoomed out, And it just, like, it'll do better if we increase the resolution, And it has easier time finding these fine grained features, Like, where the watch hands are pointing. [00:20:53] Peter Robicheaux: Nodice. And you can say, okay, well, maybe the model just doesn't know how to tell time from knowing the position of the hands. But if you actually prompt [00:21:00] it textually, it's very easy for it to tell the time. So this to me is proof that these LLMs literally cannot see the position of the watch hands and it can't see those details. [00:21:08] Peter Robicheaux: So the question is sort of why? And for you anthropic heads out there, cloud fails too. So the, the, my first pick for best paper of 2024 Envision is this MMVP paper, which tries to investigate the Why do LLMs not have the ability to see fine grained details? And so, for instance, it comes up with a lot of images like this, where you ask it a question that seems very visually apparent to us, like, which way is the school bus facing? [00:21:32] Peter Robicheaux: And it gets it wrong, and then, of course, it makes up details to support its wrong claim. And so, the process by which it finds these images is sort of contained in its hypothesis for why it can't. See these details. So it hypothesizes that models that have been initialized with, with Clip as their vision encoder, they don't have fine grained details and the, the features extracted using Clip because Clip sort of doesn't need to find these fine grained [00:22:00] details to do its job correctly, which is just to match captions and images, right? [00:22:04] Peter Robicheaux: And sort of at a high level, even if ChatGPT wasn't initialized with Clip and wasn't trained contrastively at all. The vision encoder wasn't trained contrastively at all. Still, in order to do its job of capturing the image it could do a pretty good job without actually finding the exact position of all the objects and visual features in the image, right? [00:22:21] Peter Robicheaux: So This paper finds a set of difficult images for these types of models. And the way it does it is it looks for embeddings that are similar in clip space, but far in DynaV2 space. So DynaV2 is a foundation model that was trained self supervised purely on image data. And it kind of uses like some complex student teacher framework, but essentially, and like, it patches out like certain areas of the image or like crops with certain areas of the image and tries to make sure that those have consistent representations, which is a way for it to learn very fine grained visual features. [00:22:54] Peter Robicheaux: And so if you take things that are very close in clip space and very far in DynaV2 space, you get a set of images [00:23:00] that Basically, pairs of images that are hard for a chat GPT and other big language models to distinguish. So, if you then ask it questions about this image, well, as you can see from this chart, it's going to answer the same way for both images, right? [00:23:14] Peter Robicheaux: Because to, to, from the perspective of the vision encoder, they're the same image. And so if you ask a question like, how many eyes does this animal have? It answers the same for both. And like all these other models, including Lava do the same thing, right? And so this is the benchmark that they create, which is like finding clip, like clip line pairs, which is pairs of images that are similar in clip space and creating a data set of multiple choice questions based off of those. [00:23:39] Peter Robicheaux: And so how do these models do? Well, really bad. Lava, I think, So, so, chat2BT and Jim and I do a little bit better than random guessing, but, like, half of the performance of humans who find these problems to be very easy. Lava is, interestingly, extremely negatively correlated with this dataset. It does much, much, much, much worse [00:24:00] than random guessing, which means that this process has done a very good job of identifying hard images for, for Lava, specifically. [00:24:07] Peter Robicheaux: And that's because Lava is basically not trained for very long and is initialized from Clip, and so You would expect it to do poorly on this dataset. So, one of the proposed solutions that this paper attempts is by basically saying, Okay, well if clip features aren't enough, What if we train the visual encoder of the language model also on dyno features? [00:24:27] Peter Robicheaux: And so it, it proposes two different ways of doing this. One, additively which is basically interpolating between the two features, and then one is interleaving, which is just kind of like training one on the combination of both features. So there's this really interesting trend when you do the additive mixture of features. [00:24:45] Peter Robicheaux: So zero is all clip features and one is all DynaV2 features. So. It, as you, so I think it's helpful to look at the right most chart first, which is as you increase the number of DynaV2 features, your model does worse and worse and [00:25:00] worse on the actual language modeling task. And that's because DynaV2 features were trained completely from a self supervised manner and completely in image space. [00:25:08] Peter Robicheaux: It knows nothing about text. These features aren't really compatible with these text models. And so you can train an adapter all you want, but it seems that it's in such an alien language that it's like a very hard optimization for this. These models to solve. And so that kind of supports what's happening on the left, which is that, yeah, it gets better at answering these questions if as you include more dyna V two features up to a point, but then you, when you oversaturate, it completely loses its ability to like. [00:25:36] Peter Robicheaux: Answer language and do language tasks. So you can also see with the interleaving, like they essentially double the number of tokens that are going into these models and just train on both, and it still doesn't really solve the MMVP task. It gets Lava 1. 5 above random guessing by a little bit, but it's still not close to ChachiPT or, you know, Any like human performance, obviously. [00:25:59] Peter Robicheaux: [00:26:00] So clearly this proposed solution of just using DynaV2 features directly, isn't going to work. And basically what that means is that as a as a vision foundation model, DynaV2 is going to be insufficient for language tasks, right? [00:26:14] Florence 2 (Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks) [00:26:14] Peter Robicheaux: So my next pick for best paper of 2024 would be Florence 2, which tries to solve this problem by incorporating not only This dimension of spatial hierarchy, which is to say pixel level understanding, but also in making sure to include what they call semantic granularity, which ends up, the goal is basically to have features that are sufficient for finding objects in the image, so they're, they're, they have enough pixel information, but also can be talked about and can be reasoned about. [00:26:44] Peter Robicheaux: And that's on the semantic granularity axis. So here's an example of basically three different paradigms of labeling that they do. So they, they create a big dataset. One is text, which is just captioning. And you would expect a model that's trained [00:27:00] only on captioning to have similar performance like chat2BT and like not have spatial hierarchy, not have features that are meaningful at the pixel level. [00:27:08] Peter Robicheaux: And so they add another type, which is region text pairs, which is essentially either classifying a region or You're doing object detection or doing instance segmentation on that region or captioning that region. And then they have text phrased region annotations, which is essentially a triple. And basically, not only do you have a region that you've described, you also find it's like, It's placed in a descriptive paragraph about the image, which is basically trying to introduce even more like semantic understanding of these regions. [00:27:39] Peter Robicheaux: And so like, for instance, if you're saying a woman riding on the road, right, you have to know what a woman is and what the road is and that she's on top of it. And that's, that's basically composing a bunch of objects in this visual space, but also thinking about it semantically, right? And so the way that they do this is they take basically they just dump Features from a vision encoder [00:28:00] straight into a encoder decoder transformer. [00:28:03] Peter Robicheaux: And then they train a bunch of different tasks like object detection and so on as a language task. And I think that's one of the big things that we saw in 2024 is these, these vision language models operating in, on pixel space linguistically. So they introduced a bunch of new tokens to point to locations and [00:28:22] Peter Robicheaux: So how does it work? How does it actually do? We can see if you look at the graph on the right, which is using the, the Dino, the the Dino framework your, your pre trained Florence 2 models transfer very, very well. They get 60%, 60 percent map on Cocoa, which is like approaching state of the art and they train [00:28:42] Vik Korrapati: with, and they [00:28:43] Peter Robicheaux: train with a much more more efficiently. [00:28:47] Peter Robicheaux: So they, they converge a lot faster, which both of these things are pointing to the fact that they're actually leveraging their pre trained weights effectively. So where is it falling short? So these models, I forgot to mention, Florence is a 0. 2 [00:29:00] billion and a 0. 7 billion parameter count. So they're very, very small in terms of being a language model. [00:29:05] Peter Robicheaux: And I think that. This framework, you can see saturation. So, what this graph is showing is that if you train a Florence 2 model purely on the image level and region level annotations and not including the pixel level annotations, like this, segmentation, it actually performs better as an object detector. [00:29:25] Peter Robicheaux: And what that means is that it's not able to actually learn all the visual tasks that it's trying to learn because it doesn't have enough capacity. [00:29:32] PalíGemma / PaliGemma 2 [00:29:32] Peter Robicheaux: So I'd like to see this paper explore larger model sizes, which brings us to our next big paper of 2024 or two papers. So PolyGemma came out earlier this year. [00:29:42] Peter Robicheaux: PolyGemma 2 was released, I think like a week or two ago. Oh, I forgot to mention, you can actually train You can, like, label text datasets on RoboFlow and you can train a Florence 2 model and you can actually train a PolyGemma 2 model on RoboFlow, which we got into the platform within, like, 14 hours of release, which I was really excited about. [00:29:59] Peter Robicheaux: So, anyway, so [00:30:00] PolyGemma 2, so PolyGemma is essentially doing the same thing, but instead of doing an encoder decoder, it just dumps everything into a decoder only transformer model. But it also introduced the concept of location tokens to point to objects in pixel space. PolyGemma 2, so PolyGemma uses Gemma as the language encoder, and it uses Gemma2B. [00:30:17] Peter Robicheaux: PolyGemma 2 introduces using multiple different sizes of language encoders. So, the way that they sort of get around having to do encoder decoder is they use the concept of prefix loss. Which basically means that when it's generating, tokens autoregressively, it's all those tokens in the prefix, which is like the image that it's looking at and like a description of the task that it's trying to do. [00:30:41] Peter Robicheaux: They're attending to each other fully, full attention. Which means that, you know, it can sort of. Find high level it's easier for the, the prefix to color, to color the output of the suffix and also to just find like features easily. So this is sort of [00:31:00] an example of like one of the tasks that was trained on, which is like, you describe the task in English and then you give it all these, like, You're asking for it to segment these two classes of objects, and then it finds, like, their locations using these tokens, and it finds their masks using some encoding of the masks into tokens. [00:31:24] Peter Robicheaux: And, yeah, so, one of my critiques, I guess, of PolyGemma 1, at least, is that You find that performance saturates as a pre trained model after only 300 million examples seen. So, what this graph is representing is each blue dot is a performance on some downstream task. And you can see that after seeing 300 million examples, It sort of does equally well on all of the downtrend tasks that they tried it on, which was a lot as 1 billion examples, which to me also kind of suggests a lack of capacity for this model. [00:31:58] Peter Robicheaux: PolyGemma2, [00:32:00] you can see the results on object detection. So these were transferred to to Coco. And you can see that this sort of also points to an increase in capacity being helpful to the model. You can see as. Both the resolution increases, and the parameter count of the language model increases, performance increases. [00:32:16] Peter Robicheaux: So resolution makes sense, obviously, it helps to find small images, or small objects in the image. But it also makes sense for another reason, which is that it kind of gives the model a thinking register, and it gives it more tokens to, like, process when making its predictions. But yeah, you could, you could say, oh, 43. [00:32:30] Peter Robicheaux: 6, that's not that great, like Florence 2 got 60. But this is not Training a dino or a debtor on top of this language or this image encoder. It's doing the raw language modeling task on Cocoa. So it doesn't have any of the bells and whistles. It doesn't have any of the fancy losses. It doesn't even have bipartite graph matching or anything like that. [00:32:52] Peter Robicheaux: Okay, the big result and one of the reasons that I was really excited about this paper is that they blow everything else away [00:33:00] on MMVP. I mean, 47. 3, sure, that's nowhere near human accuracy, which, again, is 94%, but for a, you know, a 2 billion language, 2 billion parameter language model to be chat2BT, that's quite the achievement. [00:33:12] Peter Robicheaux: And that sort of brings us to our final pick for paper of the year, which is AIMV2. So, AIMV2 sort of says, okay, Maybe this language model, like, maybe coming up with all these specific annotations to find features and with high fidelity and pixel space isn't actually necessary. And we can come up with an even simpler, more beautiful idea for combining you know, image tokens and pixel tokens in a way that's interfaceable for language tasks. [00:33:44] Peter Robicheaux: And this is nice because it can scale, you can come up with lots more data if you don't have to come up with all these annotations, right? So the way that it works. is it does something very, very similar to PolyGemo, where you have a vision encoder that dumps image tokens into a decoder only transformer. [00:33:59] Peter Robicheaux: But [00:34:00] the interesting thing is that it also autoregressively tries to learn the mean squared error of the image tokens. So instead of having to come up with fancy object detection or semantic, or segment, or segmentation labels, you can just try to reconstruct the image and have it learn fine grained features that way. [00:34:16] Peter Robicheaux: And it does this in kind of, I think, a beautiful way that's kind of compatible with the PolyGemma line of thinking, which is randomly sampling a prefix line of thinking Prefix length and using only this number of image tokens as the prefix. And so doing a similar thing with the causal. So the causal with prefix is the, the attention mask on the right. [00:34:35] Peter Robicheaux: So it's doing full block attention with some randomly sampled number of image tokens to then reconstruct the rest of the image and the downstream caption for that image. And so, This is the dataset that they train on. It's image or internet scale data, very high quality data created by the data filtering networks paper, essentially which is maybe The best clip data that exists. [00:34:59] Peter Robicheaux: [00:35:00] And we can see that this is finally a model that doesn't saturate. It's even at the highest parameter count, it's, it appears to be, oh, at the highest parameter account, it appears to be improving in performance with more and more samples seen. And so you can sort of think that. You know, if we just keep bumping the parameter count and increasing the example scene, which is the, the, the line of thinking for language models, then it'll keep getting better. [00:35:27] Peter Robicheaux: So how does it actually do at finding, oh, it also improves with resolution, which you would expect for a model that This is the ImageNet classification accuracy, but yeah, it does better if you increase the resolution, which means that it's actually leveraging and finding fine grained visual features. [00:35:44] Peter Robicheaux: And so how does that actually do compared to CLIP on Cocoa? Well, you can see that if you slap a transformer detection head on it, Entry now in Cocoa, it's just 60. 2, which is also within spitting distance of Soda, which means that it does a very good job of [00:36:00] finding visual features, but you could say, okay, well, wait a second. [00:36:03] Peter Robicheaux: Clip got to 59. 1, so. Like, how does this prove your claim at all? Because doesn't that mean like clip, which is known to be clip blind and do badly on MMVP, it's able to achieve a very high performance on fine, on this fine grained visual features task of object detection, well, they train on like, Tons of data. [00:36:24] Peter Robicheaux: They train on like objects, 365, Cocoa, Flickr and everything else. And so I think that this benchmark doesn't do a great job of selling how good of a pre trained model MV2 is. And we would like to see the performance on fewer data as examples and not trained to convergence on object detection. So seeing it in the real world on like a dataset, like RoboFlow 100, I think would be quite interesting. [00:36:48] Peter Robicheaux: And our, our, I guess our final, final pick for paper of 2024 would be Moondream. So introducing Vic to talk about that. [00:36:54] swyx: But overall, that was exactly what I was looking for. Like best of 2024, an amazing job. Yeah, you can, [00:37:00] if there's any other questions while Vic gets set up, like vision stuff, [00:37:07] swyx: yeah, [00:37:11] swyx: Vic, go ahead. Hi, [00:37:13] Vik Korrapati / Moondream [00:37:13] question: well, while we're getting set up, hi, over here, thanks for the really awesome talk. One of the things that's been weird and surprising is that the foundation model companies Even these MLMs, they're just like worse than RT Tether at detection still. Like, if you wanted to pay a bunch of money to auto label your detection dataset, If you gave it to OpenAI or Cloud, that would be like a big waste. [00:37:37] question: So I'm curious, just like, even Pali Gemma 2, like is worse. So, so I'm curious to hear your thoughts on like, how come, Nobody's cracked the code on like a generalist that really you know, beats a specialist model in computer vision like they have in in LLM land.[00:38:00] [00:38:01] Isaac Robinson: Okay. It's a very, very interesting question. I think it depends on the specific domain. For image classification, it's basically there. In the, in AIMv2 showed, a simple attentional probe on the pre trained features gets like 90%, which is as well as anyone does. The, the, the, the bigger question, like, why isn't it transferring to object detection, especially like real time object detection. [00:38:25] Isaac Robinson: I think, in my mind, there are two answers. One is, object detection is really, really, really the architectures are super domain specific. You know, we see these, all these super, super complicated things, and it's not super easy to, to, to build something that just transfers naturally like that, whereas image classification, you know, clip pre training transfers super, super quickly. [00:38:48] Isaac Robinson: And the other thing is, until recently, the real time object detectors didn't even really benefit from pre training. Like, you see the YOLOs that are like, essentially saturated, showing very little [00:39:00] difference with pre training improvements, with using pre trained model at all. It's not surprising, necessarily, that People aren't looking at the effects of better and better pre training on real time detection. [00:39:12] Isaac Robinson: Maybe that'll change in the next year. Does that answer your question? [00:39:17] Peter Robicheaux: Can you guys hear me? Yeah, one thing I want to add is just like, or just to summarize, basically, is that like, Until 2024, you know, we haven't really seen a combination of transformer based object detectors and fancy losses, and PolyGemma suffers from the same problem, which is basically to say that these ResNet, or like the convolutional models, they have all these, like, extreme optimizations for doing object detection, but essentially, I think it's kind of been shown now that convolution models like just don't benefit from pre training and just don't like have the level of intelligence of transformer models. [00:39:56] swyx: Awesome. Hi, [00:39:59] Vik Korrapati: can [00:40:00] you hear me? [00:40:01] swyx: Cool. I hear you. See you. Are you sharing your screen? [00:40:04] Vik Korrapati: Hi. Might have forgotten to do that. Let me do [00:40:07] swyx: that. Sorry, should have done [00:40:08] Vik Korrapati: that. [00:40:17] swyx: Here's your screen. Oh, classic. You might have to quit zoom and restart. What? It's fine. We have a capture of your screen. [00:40:34] swyx: So let's get to it. [00:40:35] Vik Korrapati: Okay, easy enough. [00:40:49] Vik Korrapati: All right. Hi, everyone. My name is Vic. I've been working on Moondream for almost a year now. Like Shawn mentioned, I just went and looked and it turns out the first version I released December [00:41:00] 29, 2023. It's been a fascinating journey. So Moonbeam started off as a tiny vision language model. Since then, we've expanded scope a little bit to also try and build some tooling, client libraries, et cetera, to help people really deploy it. [00:41:13] Vik Korrapati: Unlike traditional large models that are focused at assistant type use cases, we're laser focused on building capabilities that developers can, sorry, it's yeah, we're basically focused on building capabilities that developers can use to build vision applications that can run anywhere. So, in a lot of cases for vision more so than for text, you really care about being able to run on the edge, run in real time, etc. [00:41:40] Vik Korrapati: So That's really important. We have we have different output modalities that we support. There's query where you can ask general English questions about an image and get back human like answers. There's captioning, which a lot of our users use for generating synthetic datasets to then train diffusion models and whatnot. [00:41:57] Vik Korrapati: We've done a lot of work to minimize those sessions there. [00:42:00] So that's. Use lot. We have open vocabulary object detection built in similar to a couple of more recent models like Palagem, et cetera, where rather than having to train a dedicated model, you can just say show me soccer balls in this image or show me if there are any deer in this image, it'll detect it. [00:42:14] Vik Korrapati: More recently, earlier this month, we released pointing capability where if all you're interested in is the center of an object you can just ask it to point out where that is. This is very useful when you're doing, you know, I automation type stuff. Let's see, LA we, we have two models out right now. [00:42:33] Vik Korrapati: There's a general purpose to be para model, which runs fair. Like it's, it's it's fine if you're running on server. It's good for our local Amma desktop friends and it can run on flagship, flagship mobile phones, but it never. so much for joining us today, and we'll see you in the [00:43:00] next one. Less memory even with our not yet fully optimized inference client. [00:43:06] Vik Korrapati: So the way we built our 0. 5b model was to start with the 2 billion parameter model and prune it while doing continual training to retain performance. We, our objective during the pruning was to preserve accuracy across a broad set of benchmarks. So the way we went about it was to estimate the importance of different components of the model, like attention heads, channels MLP rows and whatnot using basically a technique based on the gradient. [00:43:37] Vik Korrapati: I'm not sure how much people want to know details. We'll be writing a paper about this, but feel free to grab me if you have more questions. Then we iteratively prune a small chunk that will minimize loss and performance retrain the model to recover performance and bring it back. The 0. 5b we released is more of a proof of concept that this is possible. [00:43:54] Vik Korrapati: I think the thing that's really exciting about this is it makes it possible for for developers to build using the 2B param [00:44:00] model and just explore, build their application, and then once they're ready to deploy figure out what exactly they need out of the model and prune those capabilities into a smaller form factor that makes sense for their deployment target. [00:44:12] Vik Korrapati: So yeah, very excited about that. Let me talk to you folks a little bit about another problem I've been working on recently, which is similar to the clocks example we've been talking about. We had a customer reach out who was talking about, like, who had a bunch of gauges out in the field. This is very common in manufacturing and oil and gas, where you have a bunch of analog devices that you need to monitor. [00:44:34] Vik Korrapati: It's expensive to. And I was like, okay, let's have humans look at that and monitor stuff and make sure that the system gets shut down when the temperature goes over 80 or something. So I was like, yeah, this seems easy enough. Happy to, happy to help you distill that. Let's, let's get it going. Turns out our model couldn't do it at all. [00:44:51] Vik Korrapati: I went and looked at other open source models to see if I could just generate a bunch of data and learn from that. Did not work either. So I was like, let's look at what the folks with [00:45:00] hundreds of billions of dollars in market cap have to offer. And yeah, that doesn't work either. My hypothesis is that like the, the way these models are trained are using a large amount of image text data scraped from the internet. [00:45:15] Vik Korrapati: And that can be biased. In the case of gauges, most gauge images aren't gauges in the wild, they're product images. Detail images like these, where it's always set to zero. It's paired with an alt text that says something like GIVTO, pressure sensor, PSI, zero to 30 or something. And so the models are fairly good at picking up those details. [00:45:35] Vik Korrapati: It'll tell you that it's a pressure gauge. It'll tell you what the brand is, but it doesn't really learn to pay attention to the needle over there. And so, yeah, that's a gap we need to address. So naturally my mind goes to like, let's use synthetic data to, Solve this problem. That works, but it's problematic because it turned out we needed millions of synthetic gauge images to get to reasonable performance. [00:45:57] Vik Korrapati: And thinking about it, reading a gauge is like [00:46:00] not a one, like it's not a zero short process in our minds, right? Like if you had to tell me the reading in Celsius for this, Real world gauge. There's two dials on there. So first you have to figure out which one you have to be paying attention to, like the inner one or the outer one. [00:46:14] Vik Korrapati: You look at the tip of the needle, you look at what labels it's between, and you count how many and do some math to figure out what that probably is. So what happens if we just add that as a Chain of thought to give the model better understanding of the different sub, to allow the model to better learn the subtasks it needs to perform to accomplish this goal. [00:46:37] Vik Korrapati: So you can see in this example, this was actually generated by the latest version of our model. It's like, okay, Celsius is the inner scale. It's between 50 and 60. There's 10 ticks. So the second tick, it's a little debatable here, like there's a weird shadow situation going on, the dial is off, so I don't know what the ground truth is, but it works okay. [00:46:57] Vik Korrapati: There's points on there that are, the points [00:47:00] over there are actually grounded. I don't know if this is easy to see, but when I click on those, there's a little red dot that moves around on the image. The model actually has to predict where this points are, I was already trying to do this with bounding boxes, but then Malmo came out with pointing capabilities. [00:47:15] Vik Korrapati: And it's like pointing is a much better paradigm to to represent this. We see pretty good results. This one's actually for clock reading. I couldn't find our chart for gauge reading at the last minute. So the light. Blue chart is with our rounded chain of thought. This measures, we have, we built a clock reading benchmark about 500 images. [00:47:37] Vik Korrapati: This measures accuracy on that. You can see it's a lot more sample efficient when you're using the chain of thought to model. Another big benefit from this approach is like, you can kind of understand how the model is. it and how it's failing. So in this example, the actual correct reading is 54 Celsius, the model output [00:48:00] 56, not too bad but you can actually go and see where it messed up. Like it got a lot of these right, except instead of saying it was on the 7th tick, it actually predicted that it was the 8th tick and that's why it went with 56. [00:48:14] Vik Korrapati: So now that you know that this. Failing in this way, you can adjust how you're doing the chain of thought to maybe say like, actually count out each tick from 40, instead of just trying to say it's the eighth tick. Or you might say like, okay, I see that there's that middle thing, I'll count from there instead of all the way from 40. [00:48:31] Vik Korrapati: So helps a ton. The other thing I'm excited about is a few short prompting or test time training with this. Like if a customer has a specific gauge that like we're seeing minor errors on, they can give us a couple of examples where like, if it's miss detecting the. Needle, they can go in and correct that in the chain of thought. [00:48:49] Vik Korrapati: And hopefully that works the next time. Now, exciting approach, we only apply it to clocks and gauges. The real question is, is it going to generalize? Probably, like, there's some science [00:49:00] from text models that when you train on a broad number of tasks, it does generalize. And I'm seeing some science with our model as well. [00:49:05] Vik Korrapati: So, in addition to the image based chain of thought stuff, I also added some spelling based chain of thought to help it understand better understand OCR, I guess. I don't understand why everyone doesn't do this, by the way. Like, it's trivial benchmark question. It's Very, very easy to nail. But I also wanted to support it for stuff like license plate, partial matching, like, hey, does any license plate in this image start with WHA or whatever? [00:49:29] Vik Korrapati: So yeah, that sort of worked. All right, that, that ends my story about the gauges. If you think about what's going on over here it's interesting that like LLMs are showing enormous. Progress in reasoning, especially with the latest set of models that we've seen, but we're not really seeing, I have a feeling that VLMs are lagging behind, as we can see with these tasks that should be very simple for a human to do [00:50:00] that are very easy to find VLMs failing at. [00:50:04] Vik Korrapati: My hypothesis on why this is the case is because On the internet, there's a ton of data that talks about how to reason. There's books about how to solve problems. There's books critiquing the books about how to solve problems. But humans are just so good at perception that we never really talk about it. [00:50:20] Vik Korrapati: Like, maybe in art books where it's like, hey, to show that that mountain is further away, you need to desaturate it a bit or whatever. But the actual data on how to, like, look at images is, isn't really present. Also, the Data we have is kind of sketched. The best source of data we have is like image all text pairs on the internet and that's pretty low quality. [00:50:40] Vik Korrapati: So yeah, I, I think our solution here is really just we need to teach them how to operate on individual tasks and figure out how to scale that out. All right. Yep. So conclusion. At Moondream we're trying to build amazing PLMs that run everywhere. Very hard problem. Much work ahead, but we're making a ton of progress and I'm really excited [00:51:00] about If anyone wants to chat about more technical details about how we're doing this or interest in collaborating, please, please hit me up. [00:51:08] Isaac Robinson: Yeah, [00:51:09] swyx: like, I always, when people say, when people say multi modality, like, you know, I always think about vision as the first among equals in all the modalities. So, I really appreciate having the experts in the room. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.latent.space/subscribe https://www.latent.space/subscribe?utm_medium=podcast&utm_campaign=CTA_2

• 0:03Speaker 00

  Welcome to Latent Space Live, our first mini-conference held at NeurIPS 2024 in Vancouver. This is Charlie, your AI co-host. When we were thinking of ways to add value to our academic conference coverage, we realised that there was a lack of good talks just recapping the best of 2024 going domain by domain. We sent out a survey to the over 900 of you who told us what you wanted, and then invited the best speakers in the latent space network to cover each field. 200 of you joined us in person throughout the day with over 2,200 watching live online. Our second featured keynote is the best of Vision 2024, with Peter Robisho and Isaac Robinson of Roboflow,

• 0:49Speaker 00

  with a special appearance from Vik Corapati of Moondream. When we did a poll of our attendees, the highest interest domain of the year was vision, and so our first port of call was our friends at Roboflow. Joseph Nelson helped us kick-start our vision coverage in episode 7 last year, and this year came back as a guest host with Nicky Ravi of Meta to cover segment Anything Too. Roboflow have consistently been the leaders in open source vision models and tooling, with their Supervision Library recently eclipsing PyTorch's Vision Library, and Roboflow Universe hosting hundreds of thousands of open source vision datasets and models.

• 1:31Speaker 00

  They have since announced a $40 million Series B led by Google Ventures, woo-hoo. This is the year that vision language models became mainstream, with every model from GPT-40 to 1, to Claude 3, to Gemini 1, and 2 to Llama, 3.2 to Mistral's Pixtral, to AI2's Pixmo, going multimodal. We asked Peter and Isaac to highlight the best work in computer vision for 2024, and they blew us away with the complete overview. As a special bonus, we also got a bonus talk from Vik Corapati at Moondream, who gave an incredible talk at this year's AI Engineer World's Fair

• 2:13Speaker 00

  on his tiny 0.5 billion parameter-pruned vision language model that absolutely slaps. As always, don't forget to check the show notes for the YouTube link to their talk as well as their slides. Watch out and take care.

• 2:29Speaker 01

  Hi, we're Isaac and Peter from Roboflow, and we're going to talk about the best papers of 2024 in computer vision.

• 2:37Speaker 01

  So for us, we define best as what made the biggest shifts in the space. And to determine that, we looked at what are some major trends that happened? And what papers most contributed to those trends. So I'm going to talk about a couple of trends. Peter's going to talk about a trend and then we're going to hand it off to Moon Dream.

• 2:58Speaker 01

  So

• 3:00Speaker 01

  the trends that I'm interested in talking about are a major transition from models that run on per image basis to models that run using the same basic ideas on video. And then also how debtors are starting to take over the real time object detection scene from YOLOs, which have been dominant for years.

• 3:22Speaker 01

  So as a highlight, we're going to talk about Sora, which from my perspective is the biggest paper of 2024, even though it came out in February.

• 3:33Speaker 01

  Is the what? Yeah, yeah. So just it's a Sora is just a post. So I'm going to fill it in with details from replication efforts, including OpenSora and related work such as Stable Diffusion video.

• 3:50Speaker 01

  And then we're also going to talk about SAM2, which applies the SAM strategy to video. And then how debtors are the improvements in 2024 to debtors that are making them a pretty good improvement to yellow based models.

• 4:05Speaker 01

  So to start this off, we're going to talk about the state of the art of video generation at the end of 2023, MagVIT. MagVIT is a discrete token, video tokenizer akin to VQGAN, but applied to video sequences. And it actually outperforms state of the art handcrafted video compression frameworks in terms of

• 4:33Speaker 01

  The bit rate versus human preference for quality and videos generated by auto regressing on these discrete tokens. Generate some pretty nice stuff, but up to like five seconds length and you know, not super detailed. And then suddenly, a few months later, we have this, which when I saw it was totally mind blowing to me. 1080p, a whole minute long. We've got light reflecting in puddles, that's reflective. Reminds me of those RTX demonstrations for next generation video games such as Cyberpunk, but with better graphics. You can see some issues in the background if you look closely, but they're kind of as with a lot of these models, the issues tend to be things that people aren't going

• 5:17Speaker 01

  to pay attention to unless they're looking for. In the same way that like six fingers on a hand, you're not going to notice is a giveaway unless you're looking for it. So yeah, as we said, Sora does not have a paper, so we're going to be filling it in with context from the rest of the computer vision scene attempting to replicate these efforts. So first step, you have an LLM caption, a huge amount of videos. This was a trick that they introduced in Dolly 3, where they train a image captioning model to just generate very high quality captions for a huge corpus and then train a diffusion model on that.

• 6:02Speaker 01

  Their Sora and replication efforts also show a bunch of other steps that are necessary for good video generation, including filtering by aesthetic score and filtering by making sure the videos have enough motion, so they're not just like, kind of, the generator's not learning to just generate static frames. So then we encode our video into a series of space time latents. Once again, Sora, very sparse in details. So the replication related works, OpenSora actually uses a MagVIT V2 itself to do this, but swapping out the discretization step with a classic VAE autoencoder framework.

• 6:48Speaker 01

  And they show that there's a lot of benefit from getting the temporal compression, which makes a lot of sense as the each sequential frames of videos have mostly redundant information.

• 7:02Speaker 01

  So by compressing against compressing in the temporal space, you allow the latent to hold a lot more semantic information while avoiding that duplicate.

• 7:16Speaker 01

  So we've got our space time latents possibly, but via there's some 3D VAE, presumably a mag VITV2.

• 7:26Speaker 01

  And then you throw it into a diffusion transformer. So I think it's personally interesting to note that OpenSora is using a mag VITV2, which originally used an autoregressive transformer decoder to model the latent space, but is now using a diffusion, diffusion transformer. So it's still a transformer happening, just the question is like, is it parameterizing this stochastic differential equation, is it parameterizing a conditional distribution via autoregression?

• 8:01Speaker 01

  It's also, it's also worth noting that most diffusion models today, the very high performers ones are switching away from the classic like DDPM, a denoising diffusion probability modeling framework to rectified flows.

• 8:16Speaker 01

  Rectified flows have a very interesting property is that as they converge, they actually get closer to being able to be sampled with a single step, which means that in practice, you can actually generate high quality samples much faster.

• 8:33Speaker 01

  Major problem of DDPM and related models for the past four years is just that they require many, many steps to generate high quality samples. So and naturally, the third step is throwing lots of compute at the problem. So I didn't, I never figured out how to manage to get this video to loop, but we see very little compute.

• 8:59Speaker 01

  Medium compute, lots of compute. This is so interesting because the original diffusion transformer paper from Facebook actually showed that, in fact, the specific hyperparameters of the transformer didn't really matter that much. What mattered was that you were just increasing the amount of compute that the model had. So I love how in the, you know, once again, little blog post, they don't even talk about like the specific hyperparameters. They say, we're using a diffusion transformer and we're just throwing more compute at it. And this is what happens. OpenSora shows similar results. The primary issue, I think, here is that no one else has 32x compute budget.

• 9:39Speaker 01

  So we end up with these, uh, we end up in the middle of the domain and most of the related work, which is still super, super cool. It's just a little disappointing considering the context. Um, so I think this is a beautiful extension of the framework that was introduced in 22 and 23 for these very high quality per image generation and then extending that to videos. It's awesome. And it's GA as of Monday, except no one can seem to get access to it because they keep shutting down the login. The next, so next paper I wanted to talk about is SAM.

• 10:22Speaker 01

  So we at Roboflow allow users to label data and train models on that data. SAM for us has saved our users 75 years of labeling time. We are the, to the best of my knowledge, the largest SAM API that exists. We also, SAM also allows us to have our users train just pure bounding box regression models and use those to generate high quality masks, um, which has the great side effect of requiring less training data to have a meaningful convergence. So most people are data limited in the real world. So anything that requires less data to get to a useful thing is that's super useful.

• 11:02Speaker 01

  Um, most of our users actually run their object, per frame object detectors on every frame in a video, or maybe not most, but many, many. And so, uh, SAM falls into this category of taking SAM2 falls into this category of taking something that really, really works and applying it to a video, which has the wonderful benefit of being plug and play with most of our, many of our users' use cases.

• 11:32Speaker 01

  We're, we're still building out a sufficiently mature pipeline to take advantage of that, but it's, it's in the works. Um, so here we've got a great example. We can click on cells and then follow them. You even notice the cell goes away and comes back and we can still, uh, keep track of it. Um, which is very challenging for, uh, existing, uh, object trackers.

• 12:01Speaker 01

  High level overview of how SAM2 works. We, uh, uh, there's a simple pipeline here where we

• 12:12Speaker 01

  can give, provide some type of prompt and it fills out the rest of the likely masks for that object throughout the rest of the video. So here we're giving a bounding box in the first frame, a set of positive negative points, or even just a simple mask. Um, I'm going to assume people are somewhat familiar with SAM. So I'm going to just give a high level overview of how SAM works. You have an image encoder that runs on every frame. SAM2 can be used on a single image, in which case the only difference between SAM2 and SAM is that image encoder, which SAM used a standard VIT.

• 12:52Speaker 01

  Um, SAM2 replaced that with a, uh, HARA hierarchical encoder, which gets approximately the same results, but leads to a six times faster inference, which is excellent, especially considering how in a, a trend of 23 was replacing the VIT with more efficient backbones.

• 13:16Speaker 01

  In the case where you're doing video segmentation, the difference is that you actually create a memory bank and you cross attend the features from the image encoder based on the memory bank. So the feature set that is created is essentially, well, I'll go more into it in a couple of slides, but we take the features from the past couple of frames plus a set of object pointers and a set of prompts and use that to generate our new masks. Then we then fuse the new masks for this frame with the image features and add that to the memory bank.

• 13:59Speaker 01

  It's, I'll say more in a minute. Just like Sam, Sam too actually uses a data engine to create its data set in that people are, they assembled a huge amount of reference data, used people to label some of it and trained the model, used the model to label more of it and asked people to refine the predictions of the model. And then ultimately, the data set is just created from the final output of the model on the reference data. It's very interesting. This paradigm is so interesting to me because it unifies a model and a data set in a way that is very unique. It seems unlikely that another model could come in

• 14:41Speaker 01

  and have such a tight relationship with the training set.

• 14:48Speaker 01

  Yeah. So brief overview of how the memory bank works. Their paper did not have a great visual, so I'm just, I'm going to fill in a bit more.

• 15:01Speaker 01

  So we take the last couple of frames from our video. And we take the last couple of frames from our video, attend that along with the set of prompts that we provided. They could come from the future. They could come from anywhere in the video as well as reference object pointers saying, by the way, here's what we've found so far. Attending to the last few frames has the interesting benefit of allowing it to model complex object motion without actually.

• 15:38Speaker 01

  By limiting the amount of frames that you attend to, you manage to keep the model running in real time. This is such an interesting topic for me because one would assume that attending to all of the frames is super essential, or having some type of summarization of all the frames is super essential for high performance. But we see in their later ablation that that actually is not the case. So here, just to make sure that there is some benchmarking happening, we just compared to some of the stuff that's came out prior. And indeed, the SAMP2 strategy does improve on the state of the art.

• 16:20Speaker 01

  This ablation deep in their dependencies was super interesting to me. We see in section C, the number of memories. One would assume that increasing the count of memories would meaningfully increase performance. And we see that it has some impact, but not the type that you'd expect. And that it meaningfully decreases speed, which justifies in my mind just having this FIFO queue of memories.

• 16:47Speaker 01

  Although in the future, I'm super interested to see a more dedicated summarization of all of the last video, not just a stacking of the last frames.

• 16:59Speaker 01

  So

• 17:01Speaker 01

  that, another extension of beautiful per-frame work into the video domain. The next trend I'm interested in talking about is this interesting, at Roboflow, we're super interested in training realtime object detectors. Those are bread and butter. And so we're doing a lot to keep track of what is actually happening in that space.

• 17:26Speaker 01

  We are finally starting to see something change. So for years, YOLOs have been the dominant way of doing realtime object detection. And we can see here that they've essentially stagnated. The, the performance between 10 and 11 is not meaningfully different. At least, you know, in this type of high-level chart. And even from the last couple of series, there's not a major change. So YOLOs have hit a plateau.

• 17:56Speaker 01

  Debtors have not. So we can look here and see the Yolo series has this plateau and then these RT debtor, LW debtor and define have meaningfully changed that plateau so that in fact, the best define models are plus 4.6 AP on Coco at the same latency.

• 18:18Speaker 01

  So three major steps to accomplish this. The first RT debtor, which is technically a 2023 paper pre-print, but published officially in 24. So I'm going to include that. I hope that's okay. Um, debtors showed that RT debtor showed that we could actually match or outspeed Yolos. Um, then LW debtor showed that pre-training is hugely effective on debtors and much less so on Yolos. And then define added the types of bells and whistles that we expect from these types, this, this arena. So the major improvements that RT debtor shows was taking the multi-scale

• 18:59Speaker 01

  features that debtors typically pass into their encoder and decoupling them into a much more efficient transformer encoder. Uh, the transformer is of course quadratic complexity. So decreasing the amount of stuff that you pass in at once is super helpful for increasing your runtime or uh, increasing your throughput. So that change basically brought us up to yellow speed. And then they do a hardcore analysis on uh, it benchmarking yellows, including the NMS step. Once you, uh, once you include the NMS in the latency calculation, you see that in fact, these debtors are

• 19:39Speaker 01

  outperforming at least this time, the, uh, the, the yellows that existed. Then LW debtor goes in and suggests that in fact, the uh, um, it, this frame, the huge boost here is from pre-training. So this is the define line. And this is the define line without pre-training. It's within range. It's still an improvement over the Yolos, but the really huge boost comes from the benefit of pre-training. When YOLO-X came out in 2021, they showed that they got much better results by having a much, much longer training time. But they found that when they did that, they actually did not benefit from pre-training.

• 20:26Speaker 01

  So you see in this graph from LW-DeeDee. In fact, YOLOs do have a real benefit from pre-training, but it goes away as we increase the training time. Then the Deeders converge much faster. LW-DeeDee transfers only 50 epochs, RT-Deeders 60 epochs. So one could assume that, in fact, the entire extra gain from pre-training is that you're not destroying your original weights by relying on this long training cycle.

• 20:57Speaker 01

  And then LW-DeeDee also shows superior performance to our favorite dataset, Roboflow-100, which means that they do better on the real world, not just on COCO.

• 21:09Speaker 01

  Then D-Fine throws all the bells and whistles at it.

• 21:13Speaker 01

  YOLO models tend to have a lot of very specific, complicated loss functions. This, uh, D-Fine brings that into the Deeder world and shows consistent improvement on a variety of Deeder-based frameworks. So bring these all together and we see that suddenly we have almost 60 AP on COCO while running in like 10 milliseconds. Huge, huge stuff. So we're spending a lot of time trying to build models that work better with less data and Deeders are clearly becoming a promising step in that direction.

• 21:50Speaker 01

  The, we're interested in seeing from the Deeders in this trend to next is Co-Deeder and the models that are currently sitting on the top of the leaderboard for large-scale inference.

• 22:05Speaker 01

  Scale really well as you switch out the backbone. We're very interested in seeing and having people publish a paper, potentially us, on what happens if you take these realtime ones and then throw a swing G at it. Like, do we have a Pareto curve that extends from the realtime domain all the way up to the super, super slow but high-performance domain? We also want to see people benchmarking an RF 100 more because that type of data is what's relevant for most users.

• 22:35Speaker 01

  And we want to see more pre-training because pre-training works now. It's super cool.

• 22:43Speaker 02

  All right, so yeah, so in that theme, one of the big things that we're focusing on is how do we get more out of our pre-trained models? And one of the lenses to look at this is through sort of this, this new requirement for like fine-grained visual details and your representations that are extracted from your foundation model. So it's sort of a hook for this. Oh, yeah, this is just a list of all the, the papers that I'm going to mention. I just wanted to make sure I said it, the actual paper, so you can find it later. Yeah, so sort of the big hook here is that I make the claim that LLMs can't see. If you go to, if you go to Claude or ChatGPT, you ask it to, to see this

• 23:27Speaker 02

  watch and tell me what time it is, it fails, right? And so you could say like, maybe, maybe the, like, this is like a very classic test of an LLM, but you could say, okay, maybe this, this image is like too zoomed out and it just like, it'll do better if we increase the resolution and it has easier time finding these fine, fine-grained features like where the watch hands are pointing. No dice. And you could say, okay, well, maybe the model just doesn't know how to tell time from knowing the position of the hands, but if you actually prompt it textually, it's very easy for it to tell the time. So this, to me, is proof that these LLMs literally cannot see the position of the watch hands and it can't see those details. So the question is sort of why? And for you anthropic heads out there, Claude fails too.

• 24:12Speaker 02

  So the my first pick for best paper of 2024 in vision is this MMVP paper, which tries to investigate why do LLMs not have the ability to see fine-grained details? And so, for instance, it it comes up with a lot of images like this where you ask it a question that seems very visually apparent to us, like which way is the school bus facing? And it gets it wrong. And then, of course, it makes up details to support its wrong claim. And so the process by which it finds these images is sort of contained in its hypothesis for why it can't see these details. So it hypothesizes that models that have been initialized with with clip as their vision encoder, they don't have fine-grained details and the

• 24:54Speaker 02

  features extracted using clip because clip sort of doesn't need to find these fine-grained details to do its job correctly, which is just to match captions and images, right? And sort of at a high level, even if chat GPT wasn't initialized with clip and wasn't trained contrastively, the vision encoder wasn't trained contrastively at all, still in order to do its job of captioning the image, it could do a pretty good job without actually finding the exact position of all the objects and visual features in the image, right? So this paper finds a set of difficult images for these types of models. And the way it does it is it looks for embeddings that are similar in clip space but far in DynaV2

• 25:35Speaker 02

  space. So DynaV2 is a foundation model that was trained self-supervised purely on image data. And it kind of uses like some complex student-teacher framework, but essentially and like it patches out like certain areas of the image or like crops of certain areas of the image and tries to make sure that those have consistent representations, which is a way for it to learn very fine-grained visual features. And so if you take things that are very close in clip space and very far in DynaV2 space, you get a set of images that Basically, pairs of images that are hard for ChatGPT and other big language models to distinguish. So if you then ask it questions about this image, well, as you can see from this chart,

• 26:17Speaker 02

  it's going to answer the same way for both images, right? Because to, from the perspectives of vision encoder, they're the same image. And so if you ask a question like, how many eyes does this animal have? It answers the same for both. And like all these other models, including Lava, do the same thing, right? And so this is the benchmark that they create, which is like finding clip, like clip line pairs, which is pairs of images that are similar in clip space, and creating a data set of multiple choice questions based off of those. And so how do these models do? Well, really bad. Lava, I think, so ChatGPT and Jim and I do a little bit better than random guessing, but like half of the performance of humans who find these problems to be very easy.

• 27:01Speaker 02

  Lava is, interestingly, extremely negatively correlated with this data set. It does much, much, much, much worse than random guessing, which means that this process has done a very good job of identifying hard images for Lava specifically. And that's because Lava is basically not trained for very long and is initialized from clip. And so you would expect it to do poorly on this data set. So one of the proposed solutions that this paper attempts is by basically saying, okay, well, if clip features aren't enough, what if we train the visual encoder of the language model also on dynamic features? And so it proposes two different ways of doing this. One, additively, which is basically interpolating between the two features.

• 27:46Speaker 02

  And then one is interleaving, which is just kind of like training one on the combination of both features. So there's this really interesting trend when you do the additive mixture of features. So zero is all clip features and one is all dynamic two features.

• 28:04Speaker 02

  It as you, so I think it's helpful to look at the rightmost chart first, which is as you increase the number of DynaV2 features, your model does worse and worse and worse on the actual language modeling task. And that's because DynaV2 features were trained completely from a self-supervised manner and completely in image space. It knows nothing about text. These features aren't really compatible with these text models. And so you can train an adapter all you want, but it seems that it's in such an alien language that it's like a very hard optimization for this, these models to solve. And so that kind of supports what's happening on the left, which is that, yeah, it gets better at answering these questions as you include more DynaV2 features up to a point. But then you, when you oversaturate,

• 28:47Speaker 02

  it completely loses its ability to like answer language and do language tasks. So you can also see with the interleaving, like they essentially double the number of tokens that are going into these models and just train on both. And it still doesn't really solve the MMVP task. It gets Lava 1.5 above random guessing by a little bit, but still not close to ChatGPT or any like human performance, obviously. So clearly, this proposed solution of just using DynaV2 features directly isn't going to work. And basically, what that means is that as a, as a vision foundation model, DynaV2 is going to be insufficient for language tasks, right?

• 29:32Speaker 02

  So my next pick for best paper of 2024 would be Florence 2, which tries to solve this problem by incorporating not only this dimension of spatial hierarchy, which is to say pixel level understanding, but also in making sure to include what they call semantic granularity, which ends up, the goal is basically to have features that are sufficient for finding objects in the image. So they're, they're. They have enough pixel information, but also can be talked about and can be reasoned about. And that's on the semantic granularity axis. So here's an example of basically three different paradigms of labeling that they do. So they create a big data set.

• 30:16Speaker 02

  One is text, which is just captioning. And you would expect a model that's trained only on captioning to have similar performance like ChatGPT and like not have spatial hierarchy, not have features that are meaningful at the pixel level. And so they add another type, which is region text pairs, which is essentially either classifying a region or you doing object detection or doing instance segmentation on that region or captioning that region. And then they have text phrase region annotations, which is essentially a triple. And basically, not only do you have a region that you've described, you also find it's like its place in a descriptive paragraph about the image, which is basically trying

• 31:01Speaker 02

  to introduce even more like semantic understanding of these regions. And so like, for instance, if you're saying a woman riding on the road, right, you have to know what a woman is and what the road is and that she's on top of it. And that's, that's basically composing a bunch of objects in this visual space, but also thinking about it semantically, right? And so the way that they do this is they take, basically they just dump features from a vision coder straight into a encoder decoder transformer. And then they train a bunch of different tasks like object detection and so on as a language task. And I think that's one of the big things that we saw in 2024 is these, these vision

• 31:45Speaker 02

  language models operating in on pixel space linguistically. So they introduce a bunch of new tokens to point to locations and in pixel space. So how does it work? How does it actually do? We can see if you look at the graph on the right, which is using the DINO framework, your pre-trained Florence 2 models transfer very, very well. They get 60% map on COCO, which is approaching state of the art. And they train with

• 32:20Speaker 02

  you're good. And they train with a much more much more efficiently. So they converge a lot faster, which both of these things are pointing to the fact that they're actually leveraging their pre-trained weights effectively.

• 32:35Speaker 02

  So where is it falling short? So these models, I forgot to mention Florence is 0.2 billion and 0.7 billion parameter count. So they're very, very small in terms of being a language model. And I think that this framework, you can see saturation. So what this graph is showing is that if you train a Florence 2 model purely on the image level and region level annotations and not including the pixel level annotations like segmentation, it actually performs better as an object detector. And what that means is that it's not able to actually learn all the visual tasks that it's trying to learn because it doesn't have enough capacity. So I'd like to see this paper explore larger model sizes,

• 33:17Speaker 02

  which brings us to our next big paper of 2024, or two papers. So PolyGemma came out earlier this year. PolyGemma 2 was released, I think, like a week or two ago. Oh, I forgot to mention you can actually train like label text datasets on Roboflow and you can train a Florence 2 model and you can actually train a PolyGemma 2 model on Roboflow, which we got into the platform within like 14 hours of release, which I was really excited about. So anyway, so PolyGemma 2 and so PolyGemma is essentially doing the same thing, but instead of doing an encoder-decoder, it just dumps everything into a decoder-only transformer model. But it also introduced the concept of location tokens to point to objects in pixel space. PolyGemma 2, so PolyGemma uses Gemma as the language encoder and it uses Gemma 2B.

• 34:02Speaker 02

  PolyGemma 2 introduces using multiple different sizes of language encoders. So the way that they sort of get around having to do encoder-decoder is they use the concept of prefix loss, which basically means that when it's generating tokens autoregressively, it's all those tokens in the prefix, which is like the image that it's looking at and like a description of the task that it's trying to do, they're attending to each other fully, full attention, which means that, you know, it can sort of find high-level, it's easier for the prefix to color the output of the suffix and also to just find like features easily.

• 34:45Speaker 02

  So this is sort of an example of like one of the tasks that was trained on, which is like you describe the task in English and then you give it all these, like you're asking for it to segment these two classes of objects and then it finds like their locations using these tokens and it finds their masks using some encoding of the masks into tokens. And yeah, so one of my critiques, I guess, of PolyGemma 1 at least, is that you find that performance saturates as a pre-trained model after only 300 million examples seen.

• 35:26Speaker 02

  So what this graph is representing is each blue dot is a performance on some downstream task. You can see that after seeing 300 million examples, it sort of does equally well on all of the downstream tasks that they tried it on, which was a lot, as 1 billion examples, which to me also kind of suggests a lack of capacity for this model. PolyGemma 2, you can see the results on object detection. So these were transferred to

• 35:57Speaker 02

  To COCO. And you can see that this sort of also points to an increase in capacity being helpful to the model. You can see as both the resolution increases and the parameter count of the language model increases, performance increases. So resolution makes sense, obviously, it helps to find small images or small objects in the image. But it also makes sense from another reason, which is that it kind of gives the model a thinking register and it gives it more tokens to like process when making its predictions.

• 36:25Speaker 02

  But yeah, you could, you could say, oh, 43.6, that's not that great. Like Florence 2 got 60, but this is not training a DINO or a DEDER on top of this language or this image encoder. It's doing the raw language modeling task on COCO. So it doesn't have any of the bells and whistles. It doesn't have any of the fancy losses. It doesn't even have bipartite graph matching or anything like that. Okay. The big result, and one of the reasons that I was really excited about this paper, is that they blow everything else away on MMVP. I mean, 47.3, sure, that's nowhere near human accuracy, which again is 94%. But for, you know, a 2 billion language, 2 billion parameter language model to be ChatGPT, that's quite the achievement.

• 37:11Speaker 02

  And that sort of brings us to our final pick for paper of the year, which is AMV2. So AMV2 sort of says, okay, maybe this language model, like maybe coming up with all these specific annotations to find features and with high fidelity in pixel space, isn't actually necessary. And we can come up with an even simpler, more beautiful idea for combining, you know, image tokens and pixel tokens in a way that's interfaceable for language tasks. And this is nice because it can scale, you can come up with lots more data if you don't have to come up with all these annotations, right? So the way that it works is it does something very, very similar to PolyGEMO,

• 37:54Speaker 02

  where you have a vision encoder that dumps image tokens into a decoder-only transformer.

• 38:01Speaker 02

  But the interesting thing is that it also auto-aggressively tries to learn the mean squared error of the image tokens. So instead of having to come up with fancy object detection or semantic or segmentation labels, you can just try to reconstruct the image and have it learn fine-grained features that way. And it does this in kind of, I think, a beautiful way that's kind of compatible with the polygema line of thinking, which is randomly sampling a prefix length and using only this number of image tokens as the prefix. And so doing a similar thing with the causal, so the causal with prefix is the attention mask on the right. So it's doing full block attention with some randomly sampled number of image tokens

• 38:43Speaker 02

  to then reconstruct the rest of the image and the downstream caption for that image. And so this is the data set that they train on. It's image or internet scale data, very high quality data created by the data filtering networks paper essentially, which is maybe the best clip data that exists. And we can see that this is finally a model that doesn't saturate. It's even at the highest parameter count, it appears to be, oh, at the highest parameter account, it appears to be improving in performance with more and more sample scene. And so you can sort of think that, you know, if we just keep bumping the parameter count

• 39:28Speaker 02

  and increasing the example scene, which is the line of thinking for language models, then it'll keep getting better. So how does it actually do at finding, oh, it also improves with resolution, which you would expect for a model that, this is the ImageNet classification accuracy, but yeah, it does better if you increase the resolution, which means that's actually leveraging and finding fine-grained visual features. And so how does it actually do compared to clip on COCO? Well, you can see that if you slap a transformer detection head on it and train it on COCO, it gets to 60.2, which is also within spitting distance of soda, which means that it does a very good job of finding visual features.

• 40:11Speaker 02

  But you could say, okay, well, wait a second, clip got to 59.1, so.

• 40:17Speaker 02

  Like, how does this prove your claim at all? Because doesn't that mean like, Clip, which is known to be Clip-blind and do badly on MMVP, it's able to achieve a very high performance on this fine-grained visual features task of object detection? Well, they train on like, tons of data. They train on like, Objects365, Coco, Flickr, and everything else. And so I think that this benchmark doesn't do a great job of seeing how good of a pre-trained model MV2 is. And we would like to see performance on fewer data as examples and not trained to convergence on object detection. So seeing it in the real world on like, a dataset like Roboflow 100, I think, would be quite interesting.

• 40:58Speaker 02

  And I guess our final, final pick for Paper of 2024 would be MoonDream. So introducing Vic to talk about that.

• 41:04Speaker 03

  But overall, that was exactly what I was looking for. Best attorney for it. Amazing job. Yeah, you can... There's anyone have questions while Vic gets set up? Like, vision stuff? Yeah? Vic, go ahead.

• 41:28Speaker 04

  Well, while we're getting set up high over here. Thanks for the really awesome talk. One of the things that's been weird and surprising is that the foundation model companies, even these MLMs, they're just like, worse than RTDeter at detection still. Like, if you wanted to pay a bunch of money to auto-label your detection dataset, if you gave it to OpenAI or Claude, that would be like, a big waste. So I'm curious, just like, even Paligema 2, like, is worse. So I'm curious to hear your thoughts on like, how come nobody's cracked the code on like,

• 42:08Speaker 04

  a generalist that really, you know, beats a specialist model in computer vision like they have in LLM land?

• 42:30Speaker 01

  It's very, very interesting question. I think it depends on the specific domain. For image classification, it's basically there in the aim V2 showed a simple attentional probe on the pre-trained features gets like 90%, which is as well as anyone does.

• 42:49Speaker 01

  The bigger question like, why isn't it transferring to object detection, especially like real-time object detection?

• 43:00Speaker 01

  I think in my mind, there are two answers. One is object detection is really, really, really. The architectures are super domain specific. You know, we see these, all these super, super complicated things and it's not super easy to, to, to build something that just transfers naturally like that, whereas image classification, you know, clip pre-training transfers super, super easily. And the other thing is, until recently, the real-time object detectors didn't even really benefit from pre-training. Like you see the YOLOs that are like essentially saturated, showing very little difference with pre-training improvements with using pre-trained model at all. It's not surprising necessarily that

• 43:43Speaker 01

  people aren't looking at the effects of better and better pre-training on real-time detection. Maybe that'll change in the next year. So to answer your question.

• 43:54Speaker 02

  Can you guys hear me? Yeah, what they want to add is just like, or just to summarize basically is that like until 2024, you know, we haven't really seen a combination of transformer-based object detectors and fancy losses. And PolyGemma suffers from the same problem, which is basically to say that

• 44:14Speaker 02

  these ResNet or like the convolutional models, they have all these like extreme optimizations for, for doing object detection, but essentially I think it's kind of been shown now that convolutional models like just don't benefit from pre-training and just don't like have the level of intelligence of transformer models.

• 44:34Speaker 03

  Awesome. Moonream.

• 44:37Speaker 05

  Hi, can you hear me?

• 44:39Speaker 03

  Cool. Hear you, see you. Are you sharing a screen?

• 44:42Speaker 05

  I might have forgotten to do that.

• 44:45Speaker 03

  Sorry.

• 44:55Speaker 03

  Share screen. Uh oh. Classic. You might have to quit Zoom and restart.

• 45:03Speaker 05

  What?

• 45:05Speaker 03

  It's fine. Yeah, it's like we have, we have a capture of your screen. I just make sure it's visible.

• 45:13Speaker 03

  So let's get to your screen.

• 45:14Speaker 05

  Okay. Easy now.

• 45:19Speaker 03

  I'll make it like we for you.

• 45:23Speaker 05

  You work up. With Zoom?

• 45:25Speaker 03

  No. Yeah, yeah.

• 45:26Speaker 05

  There we go.

• 45:27Speaker 03

  Yeah, perfect.

• 45:29Speaker 05

  All right. Hi, everyone. My name is Vic. I've been working on Moonream for almost a year now, like Sean mentioned. I just went and looked and it turns out the first version I released December twenty nine, twenty twenty three.

• 45:42Speaker 05

  It's been fascinating, generally. So Moonream started off as a tiny vision language model. Since then, we expanded scope a little bit to also try and build some tooling, client libraries, etc., to help people really deploy it. Unlike traditional large models that are focused at assistant type use cases where laser focused on building

• 46:06Speaker 05

  capabilities that developers can sorry, it's a

• 46:15Speaker 05

  yeah, we're, there's a focus on building capabilities that developers can use to build vision applications that can run anywhere. So in a lot of cases for vision, more so than for text, you really care about being able to run on the edge, run in real time, etc. So it's very important. We have we have different output modalities that we support. There's query where you can ask general English questions about an image and get back human-like answers. There's captioning, which a lot of our users use for generating synthetic data sets to then train diffusion models and whatnot. We've done a lot of work to minimize hallucinations there. So that's Um. Use a lot. We have open vocabulary object detection built in similar to a couple of more recent models like Paligema, etc., where

• 46:59Speaker 05

  rather than having to train a dedicated model, you can just say, show me soccer balls in this image, or show me there any deer in this image detected. Uh. More recently, earlier this month, we released pointing capability where if all you're interested in is the center of an object, um, you can just ask it to point out where that is. This is very useful when you're doing E O I automation type stuff.

• 47:23Speaker 05

  Um, let's see.

• 47:27Speaker 05

  La, we, we have two models out right now. There is a general purpose two B param model, which um runs fair, like it's, it's, it's fine if you're running on server. It's uh good for our local Lama desktop friends. And it can run on flagship, flagship mobile phones, but it never really fulfilled the promise of being able to run anywhere. Uh. Last week, we released a new zero point five B param model, uh, which should be seen more as a distillation target as opposed to a general purpose model. Uh, it's very good if you're running on like older mobile phones or edge devices, uses less memory even with our not yet fully optimized inference line. Um.

• 48:08Speaker 05

  So the way we built our zero point five B model was to start with the two billion parameter model um and prune it while doing continual training to retain performance. We, our objective during the pruning was to preserve accuracy across a broad set of benchmarks. So the way we went about it was to estimate the importance of different components of the model, like attention heads, channels, um, M L P rows and whatnot, um, using basically a technique based on the gradient. I'm not sure how much people want to know details. We'll be writing a paper about this, but uh, feel free to grab me if you have more questions. Uh, then we iteratively prune a small chunk that will minimize loss and performance, uh, retrain the model to recover performance and bring it back.

• 48:54Speaker 05

  Um, the zero point five B we released is more of a proof of concept that this is possible. I think the thing that's really exciting about this is it makes it possible for um.

• 49:04Speaker 05

  For developers to build using the to be param model and just explore build their application and then once they're ready to deploy, uh, figure out what exactly they need out of the model and prune those capabilities into a smaller form factor that makes sense for their deployment target. Um, so yeah, very excited about that. Let me talk to you folks a little bit about another problem I've been working on recently, which is similar to the clocks example we've been talking about. We had a customer reach out who was talking about like,

• 49:35Speaker 05

  who had a bunch of gauges out in the field. This is very common in manufacturing and oil and gas where you have a bunch of analog devices that you need to monitor. It's expensive to have humans look at that and monitor stuff and make sure that uh, the system gets shut down when the temperature goes over eighty or something. So I was like, yeah, this seems easy enough. Happy to, happy to help you distill that. Uh, let's, let's get it going. Turns out our model couldn't do it at all. Uh, I went and looked at other open source models to see if I could just generate a bunch of data and learn from that. That did not work either. So I was like, let's look at what the folks with hundreds of billions of dollars in market have to offer.

• 50:14Speaker 05

  And yeah, that doesn't work either. Um, my hypothesis is that like the, the way these models are trained are using a large amount of image text data scraped from the internet and that can be biased. In the case of gauges, most gauge images aren't gauges in the wild. They're product

• 50:34Speaker 05

  detail images like these where it's always set to zero. It's paired with an alt text that says something like G I V T O pressure sensor PSI zero to thirty or something. And so the models are fairly good at picking up those details. It'll tell you that it's a pressure gauge. It'll tell you what the brand is, but it doesn't really learn to pay attention to the needle over there.

• 50:55Speaker 05

  And so, yeah, that's a gap we need to address. So naturally, my mind goes to like, let's use synthetic data to solve this problem. Um, that works, but it's problematic because it turned out we needed millions of synthetic gauge images to get to reasonable performance. And thinking about it, reading a gauge is like not a one, like it's not a zero shot process in our minds, right? Like if you had to tell me the reading in Celsius for this real world gauge, uh, there's two dials on there. So first you have to figure out which one you have to be paying attention to, like the inner one or the outer one. Um, you look at the tip of the needle, you look at what labels it's between and you

• 51:38Speaker 05

  count how many and do some math to figure out what that probably is. So what happens if we just add that as chain of thought? Um, to give the model better understanding of the difference up. To allow the model to better learn the subtasks it needs to perform to accomplish this goal. Um, so you can see in this example, this was actually generated by the latest version of our model. Uh, it's like, okay, Celsius is the inner scale. It's between fifty and sixty, there's ten ticks.

• 52:08Speaker 05

  It's at the second tick. It's a little debatable here, like there's a weird shadow situation going on, the dial is off. So I, I don't know what the ground truth is, but

• 52:16Speaker 05

  it works. Okay. Um, there's points on there that the points over there are actually grounded. I don't know if this is easy to see, but when I click on those, there's a little red dot that moves around. On the image, the model actually has to predict where uh, those points are. I was already trying to do this with bounding boxes, but then Malmo came out with pointing capabilities and it's like pointing is a much better paradigm to uh, to represent this. We see pretty good results. This one's actually for clock reading. I couldn't find our chart for gauge reading at the last minute, so. Um, the light blue chart is with our grounded chain of thought.

• 52:59Speaker 05

  Um, this measures, we, we have, we built a clock reading benchmark about five hundred images. This measures accuracy on that. Uh, you can see it's a lot more sample efficient, uh, when you're using the chain of thought to to help the model. Um. Yep. Another big benefit from this approach is like you can kind of understand how the model is doing it and how it's failing. So in this example, the actual correct reading is fifty four Celsius, the model output fifty six, not too bad. But you can actually go and see where it messed up, like it got a lot of these right, except

• 53:42Speaker 05

  instead of saying it was on the seventh tick, it actually predicted that it was the eighth tick and that's why it went with fifty six. So now that you know that this is failing in this way, you can adjust how you're doing the chain of thought to maybe see like actually count out each tick from forty instead of just trying to say it's the eighth tick. Or you might say like, okay, I see that there's that middle thing, I'll count from there instead of all the way from forty.

• 54:07Speaker 05

  So helps a ton. The other thing I'm excited about is a few short prompting or test time training with this, like if a customer has a specific gauge that like we're seeing minor errors on, they can give us a couple of examples where like if it's mis detecting the needle, they can go in and correct that in the chain of thought and hopefully that works the next time.

• 54:30Speaker 05

  Now, exciting approach, we only apply it to clocks and gauges, the real question is, is it going to generalize? Probably like there's some signs from text models that when you train on a broad number of tasks, it does generalize and

• 54:42Speaker 05

  I'm seeing some signs with our model as well. So in addition to the image based chain of thought stuff, I also added some spelling based chain of thought to help it understand

• 54:54Speaker 05

  better than OCR, I guess. I don't understand why everyone doesn't do this, by the way, like it's trivial benchmark question, it's very, very easy to nail.

• 55:05Speaker 05

  But I also wanted to support it for stuff like license plate partial matching, like hey, does any license plate in this image start with W H A or whatever? So yeah, that sort of worked.

• 55:18Speaker 05

  All right, that that ends my story about the gauges. If you think about what's going on over here. It's interesting that like L L M's are showing enormous progress in reasoning, especially with the latest set of models that we've seen. But we're not really seeing, I I have a feeling that V L M's are lagging behind as we can see with these tasks that should be very simple for a human to do that are very easy to find. Uh, V L M's failing at. Uh, my hypothesis on why this is the case is because on the internet, there's a ton of data that talks about how to reason. There's books about how to solve problems.

• 56:01Speaker 05

  There's books creating the books about how to solve problems. But humans are just so good at perception that we never really talk about it. Like maybe in art books where it's like, hey, to show that that mountain is further away. You need to desaturate it a bit or whatever, but um. The the actual data on how to like locate images is isn't really present. Also, the data we have is kind of sketch. The best source of data we have is like image all text pairs on the internet, and that's pretty low quality. Um, so yeah, I I I think our suspicion here is really just we need to teach them how to operate on individual tasks and figure out how to scale that out. Um, all right, yep. So conclusion.

• 56:44Speaker 05

  Uh, at Moonbeam, we're trying to build amazing B L M's that run everywhere. Very hard problem, much work ahead, but uh, we're making a ton of progress and really excited about. Um, if anyone wants to chat about more um, technical details about how we're doing this or interested in collaborating, please, please hit me up.

• 57:04Speaker 03

  Yeah, like I always, when people say, when people say multimodality, like, you know, he's talking about vision as the first amount equals in all the modalities. So I really appreciate having the experts in the room.