ReLAM: Learning Anticipation Model for Rewarding Visual Robotic Manipulation

• URL: http://arxiv.org/abs/2509.22402v1
• Date: Fri, 26 Sep 2025 14:28:42 GMT
• Title: ReLAM: Learning Anticipation Model for Rewarding Visual Robotic Manipulation
• Authors: Nan Tang, Jing-Cheng Pang, Guanlin Li, Chao Qian, Yang Yu,
• Abstract summary: Reward design remains a critical bottleneck in visual reinforcement learning for robotic manipulation.<n>In this study, we propose a method that implicitly infers spatial distances through keypoints extracted from images.<n>We introduce Reward Learning with Anticipation Model (ReLAM), a novel framework that automatically generates dense, structured rewards from action-free video demonstrations.
• Score: 25.115056940401164
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Reward design remains a critical bottleneck in visual reinforcement learning (RL) for robotic manipulation. In simulated environments, rewards are conventionally designed based on the distance to a target position. However, such precise positional information is often unavailable in real-world visual settings due to sensory and perceptual limitations. In this study, we propose a method that implicitly infers spatial distances through keypoints extracted from images. Building on this, we introduce Reward Learning with Anticipation Model (ReLAM), a novel framework that automatically generates dense, structured rewards from action-free video demonstrations. ReLAM first learns an anticipation model that serves as a planner and proposes intermediate keypoint-based subgoals on the optimal path to the final goal, creating a structured learning curriculum directly aligned with the task's geometric objectives. Based on the anticipated subgoals, a continuous reward signal is provided to train a low-level, goal-conditioned policy under the hierarchical reinforcement learning (HRL) framework with provable sub-optimality bound. Extensive experiments on complex, long-horizon manipulation tasks show that ReLAM significantly accelerates learning and achieves superior performance compared to state-of-the-art methods.

Related papers

• Self-Correcting VLA: Online Action Refinement via Sparse World Imagination [55.982504915794514]
  We propose Self-Correcting VLA (SC-VLA), which achieve self-improvement by intrinsically guiding action refinement through sparse imagination.<n>SC-VLA achieve state-of-the-art performance, yielding the highest task throughput with 16% fewer steps and a 9% higher success rate than the best-performing baselines.
  arXiv  Detail & Related papers  (2026-02-25T06:58:06Z)
• ReWorld: Multi-Dimensional Reward Modeling for Embodied World Models [27.729654985554372]
  ReWorld is a framework aimed to employ reinforcement learning to align the video-based embodied world models with physical realism, task completion capability, embodiment plausibility and visual quality.<n>We show that ReWorld significantly improves the physical fidelity, logical coherence, embodiment and visual quality of generated rollouts, outperforming previous methods.
  arXiv  Detail & Related papers  (2026-01-18T14:27:10Z)
• Reinforcement Learning with Inverse Rewards for World Model Post-training [29.19830208692156]
  We propose Reinforcement Learning with Inverse Rewards to improve action-following in video world models.<n>RLIR derives verifiable reward signals by recovering input actions from generated videos using an Inverse Dynamics Model.
  arXiv  Detail & Related papers  (2025-09-28T16:27:47Z)
• Continual Visual Reinforcement Learning with A Life-Long World Model [55.05017177980985]
  We present a new continual learning approach for visual dynamics modeling.<n>We first introduce the life-long world model, which learns task-specific latent dynamics.<n>Then, we address the value estimation challenge for previous tasks with the exploratory-conservative behavior learning approach.
  arXiv  Detail & Related papers  (2023-03-12T05:08:03Z)
• Discrete Factorial Representations as an Abstraction for Goal Conditioned Reinforcement Learning [99.38163119531745]
  We show that applying a discretizing bottleneck can improve performance in goal-conditioned RL setups. We experimentally prove the expected return on out-of-distribution goals, while still allowing for specifying goals with expressive structure.
  arXiv  Detail & Related papers  (2022-11-01T03:31:43Z)
• Simplifying Model-based RL: Learning Representations, Latent-space Models, and Policies with One Objective [142.36200080384145]
  We propose a single objective which jointly optimize a latent-space model and policy to achieve high returns while remaining self-consistent. We demonstrate that the resulting algorithm matches or improves the sample-efficiency of the best prior model-based and model-free RL methods.
  arXiv  Detail & Related papers  (2022-09-18T03:51:58Z)
• Texture-guided Saliency Distilling for Unsupervised Salient Object Detection [67.10779270290305]
  We propose a novel USOD method to mine rich and accurate saliency knowledge from both easy and hard samples. Our method achieves state-of-the-art USOD performance on RGB, RGB-D, RGB-T, and video SOD benchmarks.
  arXiv  Detail & Related papers  (2022-07-13T02:01:07Z)
• Imaginary Hindsight Experience Replay: Curious Model-based Learning for Sparse Reward Tasks [9.078290260836706]
  We propose a model-based method tailored for sparse-reward tasks that foregoes the need for complicated reward engineering. This approach, termed Imaginary Hindsight Experience Replay, minimises real-world interactions by incorporating imaginary data into policy updates. Upon evaluation, this approach provides an order of magnitude increase in data-efficiency on average versus the state-of-the-art model-free method in the benchmark OpenAI Gym Fetch Robotics tasks.
  arXiv  Detail & Related papers  (2021-10-05T23:38:31Z)
• Model-Based Reinforcement Learning via Latent-Space Collocation [110.04005442935828]
  We argue that it is easier to solve long-horizon tasks by planning sequences of states rather than just actions. We adapt the idea of collocation, which has shown good results on long-horizon tasks in optimal control literature, to the image-based setting by utilizing learned latent state space models.
  arXiv  Detail & Related papers  (2021-06-24T17:59:18Z)
• Learning Long-term Visual Dynamics with Region Proposal Interaction Networks [75.06423516419862]
  We build object representations that can capture inter-object and object-environment interactions over a long-range. Thanks to the simple yet effective object representation, our approach outperforms prior methods by a significant margin.
  arXiv  Detail & Related papers  (2020-08-05T17:48:00Z)
• Hindsight Expectation Maximization for Goal-conditioned Reinforcement Learning [26.631740480100724]
  We propose a graphical model framework for goal-conditioned RL, with an EM algorithm that operates on the lower bound of the RL objective. The E-step provides a natural interpretation of how 'learning in hindsight' techniques, such as HER, to handle extremely sparse goal-conditioned rewards. The M-step reduces policy optimization to supervised learning updates, which greatly stabilizes end-to-end training on high-dimensional inputs such as images.
  arXiv  Detail & Related papers  (2020-06-13T03:25:31Z)
• Learning View and Target Invariant Visual Servoing for Navigation [9.873635079670093]
  We learn viewpoint invariant and target invariant visual servoing for local mobile robot navigation. We train deep convolutional network controller to reach the desired goal.
  arXiv  Detail & Related papers  (2020-03-04T20:36:43Z)

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.