search

MarkFzp / act-plus-plus

Imitation learning algorithms with Co-training for Mobile ALOHA: ACT, Diffusion Policy, VINN
https://mobile-aloha.github.io/
MIT License
2.86k stars 525 forks source link

Paper explain #19

Open Doan-IT opened 6 months ago

Doan-IT commented 6 months ago

Thank you for sharing such a great project. Refer to the paper "https://mobile-aloha.github.io/resources/mobile-aloha.pdf" I have some questions, hoping for your help:

  1. What is the definition of "Co-train"? Source corresponds to "co-train"
  2. "VINN + Chunking" is "CNNMLP" in the source code, right?

Thanks a lot.

kuaile258890 commented 6 months ago
  1. In the source code, VINN + Chunking should not be confused with CNNMLP. They represent different baselines. For more details, refer to the original ALOHA paper (not to be mistaken with mobile-ALOHA).
kuaile258890 commented 6 months ago

In the original ALOHA paper, BC-ConvMLP (also known as CNNMLP) is identified as the simplest yet most widely used baseline [69, 26]. It processes the current image observations with a convolutional network. The output features of this network are then concatenated with the joint positions to predict the action."