Jacob Walker

João Carreira, Dilara Gokay, Michael King et al.

Scaling has not yet been convincingly demonstrated for pure self-supervised learning from video. However, prior work has focused evaluations on semantic-related tasks $\unicode{x2013}$ action classification, ImageNet classification, etc. In this paper we focus on evaluating self-supervised learning on non-semantic vision tasks that are more spatial (3D) and temporal (+1D = 4D), such as camera pose estimation, point and object tracking, and depth estimation. We show that by learning from very large video datasets, masked auto-encoding (MAE) with transformer video models actually scales, consistently improving performance on these 4D tasks, as model size increases from 20M all the way to the largest by far reported self-supervised video model $\unicode{x2013}$ 22B parameters. Rigorous apples-to-apples comparison with many recent image and video models demonstrates the benefits of scaling 4D representations. Pretrained models are available at https://github.com/google-deepmind/representations4d .

Jacob C Walker, Pedro Vélez, Luisa Polania Cabrera et al.

Forecasting what will happen next is a critical skill for general-purpose systems that plan or act in the world at different levels of abstraction. In this paper, we identify a strong correlation between a vision model's perceptual ability and its generalist forecasting performance over short time horizons. This trend holds across a diverse set of pretrained models-including those trained generatively-and across multiple levels of abstraction, from raw pixels to depth, point tracks, and object motion. The result is made possible by a novel generalist forecasting framework that operates on any frozen vision backbone: we train latent diffusion models to forecast future features in the frozen representation space, which are then decoded via lightweight, task-specific readouts. To enable consistent evaluation across tasks, we introduce distributional metrics that compare distributional properties directly in the space of downstream tasks and apply this framework to nine models and four tasks. Our results highlight the value of bridging representation learning and generative modeling for temporally grounded video understanding.

Ankesh Anand, Jacob Walker, Yazhe Li et al.

One of the key promises of model-based reinforcement learning is the ability to generalize using an internal model of the world to make predictions in novel environments and tasks. However, the generalization ability of model-based agents is not well understood because existing work has focused on model-free agents when benchmarking generalization. Here, we explicitly measure the generalization ability of model-based agents in comparison to their model-free counterparts. We focus our analysis on MuZero (Schrittwieser et al., 2020), a powerful model-based agent, and evaluate its performance on both procedural and task generalization. We identify three factors of procedural generalization -- planning, self-supervised representation learning, and procedural data diversity -- and show that by combining these techniques, we achieve state-of-the art generalization performance and data efficiency on Procgen (Cobbe et al., 2019). However, we find that these factors do not always provide the same benefits for the task generalization benchmarks in Meta-World (Yu et al., 2019), indicating that transfer remains a challenge and may require different approaches than procedural generalization. Overall, we suggest that building generalizable agents requires moving beyond the single-task, model-free paradigm and towards self-supervised model-based agents that are trained in rich, procedural, multi-task environments.

Andrea Banino, Adrià Puidomenech Badia, Jacob Walker et al.

Many reinforcement learning (RL) agents require a large amount of experience to solve tasks. We propose Contrastive BERT for RL (CoBERL), an agent that combines a new contrastive loss and a hybrid LSTM-transformer architecture to tackle the challenge of improving data efficiency. CoBERL enables efficient, robust learning from pixels across a wide range of domains. We use bidirectional masked prediction in combination with a generalization of recent contrastive methods to learn better representations for transformers in RL, without the need of hand engineered data augmentations. We find that CoBERL consistently improves performance across the full Atari suite, a set of control tasks and a challenging 3D environment.

Jacob Walker, Ali Razavi, Aäron van den Oord

In recent years, the task of video prediction-forecasting future video given past video frames-has attracted attention in the research community. In this paper we propose a novel approach to this problem with Vector Quantized Variational AutoEncoders (VQ-VAE). With VQ-VAE we compress high-resolution videos into a hierarchical set of multi-scale discrete latent variables. Compared to pixels, this compressed latent space has dramatically reduced dimensionality, allowing us to apply scalable autoregressive generative models to predict video. In contrast to previous work that has largely emphasized highly constrained datasets, we focus on very diverse, large-scale datasets such as Kinetics-600. We predict video at a higher resolution on unconstrained videos, 256x256, than any other previous method to our knowledge. We further validate our approach against prior work via a crowdsourced human evaluation.

Jovana Mitrovic, Brian McWilliams, Jacob Walker et al.

Self-supervised learning has emerged as a strategy to reduce the reliance on costly supervised signal by pretraining representations only using unlabeled data. These methods combine heuristic proxy classification tasks with data augmentations and have achieved significant success, but our theoretical understanding of this success remains limited. In this paper we analyze self-supervised representation learning using a causal framework. We show how data augmentations can be more effectively utilized through explicit invariance constraints on the proxy classifiers employed during pretraining. Based on this, we propose a novel self-supervised objective, Representation Learning via Invariant Causal Mechanisms (ReLIC), that enforces invariant prediction of proxy targets across augmentations through an invariance regularizer which yields improved generalization guarantees. Further, using causality we generalize contrastive learning, a particular kind of self-supervised method, and provide an alternative theoretical explanation for the success of these methods. Empirically, ReLIC significantly outperforms competing methods in terms of robustness and out-of-distribution generalization on ImageNet, while also significantly outperforming these methods on Atari achieving above human-level performance on $51$ out of $57$ games.

Jacob Walker, Kenneth Marino, Abhinav Gupta et al.

Current approaches in video forecasting attempt to generate videos directly in pixel space using Generative Adversarial Networks (GANs) or Variational Autoencoders (VAEs). However, since these approaches try to model all the structure and scene dynamics at once, in unconstrained settings they often generate uninterpretable results. Our insight is to model the forecasting problem at a higher level of abstraction. Specifically, we exploit human pose detectors as a free source of supervision and break the video forecasting problem into two discrete steps. First we explicitly model the high level structure of active objects in the scene---humans---and use a VAE to model the possible future movements of humans in the pose space. We then use the future poses generated as conditional information to a GAN to predict the future frames of the video in pixel space. By using the structured space of pose as an intermediate representation, we sidestep the problems that GANs have in generating video pixels directly. We show through quantitative and qualitative evaluation that our method outperforms state-of-the-art methods for video prediction.

Jacob Walker, Carl Doersch, Abhinav Gupta et al.

In a given scene, humans can often easily predict a set of immediate future events that might happen. However, generalized pixel-level anticipation in computer vision systems is difficult because machine learning struggles with the ambiguity inherent in predicting the future. In this paper, we focus on predicting the dense trajectory of pixels in a scene, specifically what will move in the scene, where it will travel, and how it will deform over the course of one second. We propose a conditional variational autoencoder as a solution to this problem. In this framework, direct inference from the image shapes the distribution of possible trajectories, while latent variables encode any necessary information that is not available in the image. We show that our method is able to successfully predict events in a wide variety of scenes and can produce multiple different predictions when the future is ambiguous. Our algorithm is trained on thousands of diverse, realistic videos and requires absolutely no human labeling. In addition to non-semantic action prediction, we find that our method learns a representation that is applicable to semantic vision tasks.

Jacob Walker, Abhinav Gupta, Martial Hebert

Given a scene, what is going to move, and in what direction will it move? Such a question could be considered a non-semantic form of action prediction. In this work, we present a convolutional neural network (CNN) based approach for motion prediction. Given a static image, this CNN predicts the future motion of each and every pixel in the image in terms of optical flow. Our CNN model leverages the data in tens of thousands of realistic videos to train our model. Our method relies on absolutely no human labeling and is able to predict motion based on the context of the scene. Because our CNN model makes no assumptions about the underlying scene, it can predict future optical flow on a diverse set of scenarios. We outperform all previous approaches by large margins.