DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning
Abstract
Multimodal pretraining has emerged as an effective strategy for the trinity of goals of representation learning in autonomous robots: 1) extracting both local and global task progression information; 2) enforcing temporal consistency of visual representation; 3) capturing trajectory-level language grounding. Most existing methods approach these via separate objectives, which often reach sub-optimal solutions. In this paper, we propose a universal unified objective that can simultaneously extract meaningful task progression information from image sequences and seamlessly align them with language instructions. We discover that via implicit preferences, where a visual trajectory inherently aligns better with its corresponding language instruction than mismatched pairs, the popular Bradley-Terry model can transform into representation learning through proper reward reparameterizations. The resulted framework, DecisionNCE, mirrors an InfoNCE-style objective but is distinctively tailored for decision-making tasks, providing an embodied representation learning framework that elegantly extracts both local and global task progression features, with temporal consistency enforced through implicit time contrastive learning, while ensuring trajectory-level instruction grounding via multimodal joint encoding. Evaluation on both simulated and real robots demonstrates that DecisionNCE effectively facilitates diverse downstream policy learning tasks, offering a versatile solution for unified representation and reward learning.
Jianxiong Li
PhD Student
Jinliang Zheng
PhD Student
Yinan Zheng
PhD Student
Liyuan Mao
Undergraduate student at Shanghai Jiao Tong University
My research interests include reinforcement learning, especially offline reinforcement learning and imitation learning.
Xiao Hu
Research Intern
