Provide an RGB-D image to generate an affordence map directly from your pre-trained model

andyzeng / arc-robot-vision

MIT-Princeton Vision Toolbox for Robotic Pick-and-Place at the Amazon Robotics Challenge 2017 - Robotic Grasping and One-shot Recognition of Novel Objects with Deep Learning.

http://arc.cs.princeton.edu

Apache License 2.0

295 stars 97 forks source link

Provide an RGB-D image to generate an affordence map directly from your pre-trained model #11

Open wyf2015fei opened 2 years ago

wyf2015fei commented 2 years ago

Are pre-trained models available?

HanwenCao commented 2 years ago

Let me help the author to answer your question. Yes.

wyf2015fei commented 2 years ago

HanwenCao

I mean, I provide my own rgb-d image and directly generate the Heatmap, but I can't find the corresponding code. What is provided in the library can only generate the Heatmap in their dataset.