UniBYD: A Unified Framework for Learning Robotic Manipulation Across Embodiments Beyond Imitation of Human Demonstrations

2025-12-12

AI Summary

In embodied intelligence, morphological discrepancies between robotic and human hands severely limit imitation learning performance. This paper proposes UniBYD, a framework that transcends the conventional “copycat imitation” paradigm to enable robot-centric autonomous manipulation policy learning. Its core contributions are: (1) a Unified Morphological Representation (UMR) that captures kinematic commonalities across diverse hand geometries; and (2) a dynamic PPO algorithm integrating annealed reward scheduling with a hybrid Markovian shadow engine, enhancing policy generalization and action precision. Evaluated on the newly introduced UniManip benchmark—featuring multi-morphology manipulation tasks—UniBYD achieves a 67.90% higher success rate than state-of-the-art methods. It is the first approach to enable efficient, robust policy transfer across heterogeneous robotic hands and faithful reproduction of fine-grained manipulation behaviors.

Technology Category

• Intelligent Robots: Manipulation
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

In embodied intelligence, the embodiment gap between robotic and human hands brings significant challenges for learning from human demonstrations. Although some studies have attempted to bridge this gap using reinforcement learning, they remain confined to merely reproducing human manipulation, resulting in limited task performance. In this paper, we propose UniBYD, a unified framework that uses a dynamic reinforcement learning algorithm to discover manipulation policies aligned with the robot's physical characteristics. To enable consistent modeling across diverse robotic hand morphologies, UniBYD incorporates a unified morphological representation (UMR). Building on UMR, we design a dynamic PPO with an annealed reward schedule, enabling reinforcement learning to transition from imitation of human demonstrations to explore policies adapted to diverse robotic morphologies better, thereby going beyond mere imitation of human hands. To address the frequent failures of learning human priors in the early training stage, we design a hybrid Markov-based shadow engine that enables reinforcement learning to imitate human manipulations in a fine-grained manner. To evaluate UniBYD comprehensively, we propose UniManip, the first benchmark encompassing robotic manipulation tasks spanning multiple hand morphologies. Experiments demonstrate a 67.90% improvement in success rate over the current state-of-the-art. Upon acceptance of the paper, we will release our code and benchmark at https://github.com/zhanheng-creator/UniBYD.

Problem

• Bridging the embodiment gap between robotic and human hands for manipulation learning.
• Moving beyond imitation to discover policies aligned with robot physical characteristics.
• Enabling consistent modeling across diverse robotic hand morphologies.

Innovation

• Dynamic reinforcement learning algorithm for robot-specific policies
• Unified morphological representation for diverse robotic hands
• Hybrid Markov-based shadow engine for fine-grained human imitation