RESTA: Time-Scaling Augmentation for Data-Efficient Robot Imitation Learning

Key points

• Robot manipulation policies trained through imitation learning commonly rely on costly demonstrations recorded in one task direction and at one temporal resolution.1
• Trajectory-reversal augmentation reverses and relabels demonstrations between paired tasks with opposite objectives, reusing recorded rollouts without additional environment interaction.1
• Experiments across 16 paired bimanual manipulation tasks show that trajectory reversal is particularly effective in the low-data regime and that temporal-stride augmentation improves success under randomized temporal schedules.1
• These results show that RESTA can improve data efficiency in robot imitation learning.1

Sources (1)

• [1]RESTA: Time-Scaling Augmentation for Data-Efficient Robot Imitation Learning
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 11:07 AM

  We propose Reversal and Stride-based Trajectory Augmentation (RESTA), which exploits time-scaling data augmentation to improve data efficiency through trajectory reversal and temporal resampling.

  Robot manipulation policies trained through imitation learning commonly rely on costly demonstrations recorded in one task direction and at one temporal resolution.

Extractive summary: sentences quoted from the sources.