Official repository of SegDAC: Improving Visual Reinforcement Learning by Extracting Dynamic Object-Centric Representations from Pretrained Vision Models
machine-learning reinforcement-learning deep-learning robotics manipulation visual-reinforcement-learning visual-generalization maniskill3 segdac segdac-pytorch
-
Updated
Sep 25, 2025 - Jupyter Notebook