CoMo: Learning Continuous Latent Motion from Internet Videos for Scalable Robot Learning

• URL: http://arxiv.org/abs/2505.17006v1
• Date: Thu, 22 May 2025 17:58:27 GMT
• Title: CoMo: Learning Continuous Latent Motion from Internet Videos for Scalable Robot Learning
• Authors: Jiange Yang, Yansong Shi, Haoyi Zhu, Mingyu Liu, Kaijing Ma, Yating Wang, Gangshan Wu, Tong He, Limin Wang,
• Abstract summary: CoMo aims to learn more informative continuous motion representations from diverse, internet-scale videos.<n>We introduce two new metrics for more robustly and affordably evaluating motion and guiding motion learning methods.<n>CoMo exhibits strong zero-shot generalization, enabling it to generate continuous pseudo actions for previously unseen video domains.
• Score: 47.195002937893115
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Learning latent motion from Internet videos is crucial for building generalist robots. However, existing discrete latent action methods suffer from information loss and struggle with complex and fine-grained dynamics. We propose CoMo, which aims to learn more informative continuous motion representations from diverse, internet-scale videos. CoMo employs a early temporal feature difference mechanism to prevent model collapse and suppress static appearance noise, effectively discouraging shortcut learning problem. Furthermore, guided by the information bottleneck principle, we constrain the latent motion embedding dimensionality to achieve a better balance between retaining sufficient action-relevant information and minimizing the inclusion of action-irrelevant appearance noise. Additionally, we also introduce two new metrics for more robustly and affordably evaluating motion and guiding motion learning methods development: (i) the linear probing MSE of action prediction, and (ii) the cosine similarity between past-to-current and future-to-current motion embeddings. Critically, CoMo exhibits strong zero-shot generalization, enabling it to generate continuous pseudo actions for previously unseen video domains. This capability facilitates unified policy joint learning using pseudo actions derived from various action-less video datasets (such as cross-embodiment videos and, notably, human demonstration videos), potentially augmented with limited labeled robot data. Extensive experiments show that policies co-trained with CoMo pseudo actions achieve superior performance with both diffusion and autoregressive architectures in simulated and real-world settings.

Related papers

• AMPLIFY: Actionless Motion Priors for Robot Learning from Videos [29.799207502031496]
  We introduce AMPLIFY, a novel framework that leverages large-scale video data.<n>We train a forward dynamics model on abundant action-free videos and an inverse dynamics model on a limited set of action-labeled examples.<n>In downstream policy learning, our dynamics predictions enable a 1.2-2.2x improvement in low-data regimes, a 1.4x average improvement by learning from action-free human videos, and the first generalization to LIBERO tasks from zero in-distribution action data.
  arXiv  Detail & Related papers  (2025-06-17T05:31:42Z)
• MoSiC: Optimal-Transport Motion Trajectory for Dense Self-Supervised Learning [66.53533434848369]
  We propose a motion-guided self-learning framework that learns densely consistent representations.<n>We improve state-of-the-art by 1% to 6% on six image and video datasets and four evaluation benchmarks.
  arXiv  Detail & Related papers  (2025-06-10T11:20:32Z)
• Multi-Timescale Motion-Decoupled Spiking Transformer for Audio-Visual Zero-Shot Learning [73.7808110878037]
  This paper proposes a novel dual-stream Multi-Timescale Motion-Decoupled Spiking Transformer (MDST++)<n>By converting RGB images to events, our method captures motion information more accurately and mitigates background scene biases.<n>Our experiments validate the effectiveness of MDST++, demonstrating their consistent superiority over state-of-the-art methods on mainstream benchmarks.
  arXiv  Detail & Related papers  (2025-05-26T13:06:01Z)
• GENMO: A GENeralist Model for Human MOtion [64.16188966024542]
  We present GENMO, a unified Generalist Model for Human Motion that bridges motion estimation and generation in a single framework.<n>Our key insight is to reformulate motion estimation as constrained motion generation, where the output motion must precisely satisfy observed conditioning signals.<n>Our novel architecture handles variable-length motions and mixed multimodal conditions (text, audio, video) at different time intervals, offering flexible control.
  arXiv  Detail & Related papers  (2025-05-02T17:59:55Z)
• Unified World Models: Coupling Video and Action Diffusion for Pretraining on Large Robotic Datasets [7.667819384855409]
  We present Unified World Models (UWM), a framework that allows for leveraging both video and action data for policy learning.<n>By simply controlling each diffusion timestep, UWM can flexibly represent a policy, a forward dynamics, an inverse dynamics, and a video generator.<n>Our results suggest that UWM offers a promising step toward harnessing large, heterogeneous datasets for scalable robot learning.
  arXiv  Detail & Related papers  (2025-04-03T17:38:59Z)
• Self-Supervised Learning of Motion Concepts by Optimizing Counterfactuals [13.202236467650033]
  Estimating motion in videos is an essential computer vision problem with many downstream applications.<n>We develop Opt-CWM, a self-supervised technique for flow and occlusion estimation from a pre-trained next-frame prediction model.<n>We achieve state-of-the-art performance for motion estimation on real-world videos while requiring no labeled data.
  arXiv  Detail & Related papers  (2025-03-25T17:58:52Z)
• A Self-supervised Motion Representation for Portrait Video Generation [19.56640370303683]
  We propose Semantic Latent Motion (SeMo), a compact and expressive motion representation.<n>Our approach achieves both high-quality visual results and efficient inference.<n>Our approach surpasses state-of-the-art models with an 81% win rate in realism.
  arXiv  Detail & Related papers  (2025-03-13T06:43:21Z)
• Moto: Latent Motion Token as the Bridging Language for Learning Robot Manipulation from Videos [64.48857272250446]
  We introduce Moto, which converts video content into latent Motion Token sequences by a Latent Motion Tokenizer.<n>We pre-train Moto-GPT through motion token autoregression, enabling it to capture diverse visual motion knowledge.<n>To transfer learned motion priors to real robot actions, we implement a co-fine-tuning strategy that seamlessly bridges latent motion token prediction and real robot control.
  arXiv  Detail & Related papers  (2024-12-05T18:57:04Z)
• Any-point Trajectory Modeling for Policy Learning [64.23861308947852]
  We introduce Any-point Trajectory Modeling (ATM) to predict future trajectories of arbitrary points within a video frame. ATM outperforms strong video pre-training baselines by 80% on average. We show effective transfer learning of manipulation skills from human videos and videos from a different robot morphology.
  arXiv  Detail & Related papers  (2023-12-28T23:34:43Z)
• Learning Variational Motion Prior for Video-based Motion Capture [31.79649766268877]
  We present a novel variational motion prior (VMP) learning approach for video-based motion capture. Our framework can effectively reduce temporal jittering and failure modes in frame-wise pose estimation. Experiments over both public datasets and in-the-wild videos have demonstrated the efficacy and generalization capability of our framework.
  arXiv  Detail & Related papers  (2022-10-27T02:45:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.