2023/04/10 by Yingdong Hu, Renhao Wang, Hu, Yingdong +5 · 2 citations
Computer Science · Engineering · Neuroscience · #Computer Vision and Pattern Recognition (cs.CV) #EEG and Brain-Computer Interfaces #FOS: Computer and information sciences #Muscle activation and electromyography studies #Reinforcement Learning in Robotics #Robotics (cs.RO)
paper · pdf · doi:10.48550/arxiv.2304.04591
openalex publication_date 2023/04/10 · openalex created_date 2023/04/12 · openalex updated_date 2026/07/28
In recent years, increasing attention has been directed to leveraging pre-trained vision models for motor control. While existing works mainly emphasize the importance of this pre-training phase, the arguably equally important role played by downstream policy learning during control-specific fine-tuning is often neglected. It thus remains unclear if pre-trained vision models are consistent in their effectiveness under different control policies. To bridge this gap in understanding, we conduct a comprehensive study on 14 pre-trained vision models using 3 distinct classes of policy learning methods, including reinforcement learning (RL), imitation learning through behavior cloning (BC), and imitation learning with a visual reward function (VRF). Our study yields a series of intriguing results, including the discovery that the effectiveness of pre-training is highly dependent on the choice of the downstream policy learning algorithm. We show that conventionally accepted evaluation based on RL methods is highly variable and therefore unreliable, and further advocate for using more robust methods like VRF and BC. To facilitate more universal evaluations of pre-trained models and their policy learning methods in the future, we also release a benchmark of 21 tasks across 3 different environments alongside our work.