sudeepdasari / data4robotics Goto Github PK

at models/resnet.py line 9, it writes:

from r3m import load_r3m

So, I think maybe the command in README.md:

conda env create -f env.yml
conda activate data4robotics
pip install git+https://github.com/AGI-Labs/robobuf.git
pip install -e ./

should add another line:

pip install git+https://github.com/facebookresearch/r3m.git

Hi,

in the paper "HRP: Human Affordances for Robotic Pre-Training" you referred to the Appendix a few times. I wanted to check it out, but I couldn't find an upload of the paper where the Appendix is included.

Is it possible to upload it somewhere? Thanks!

@SudeepDasari Dear authors of the paper, thank you so much for your great work and the release of the code. I wonder if it is possible to use the pre-trained model to predict affordance from the image like in the VRB paper. Could you also provide the code and weight for that? Thank you so much!

Hi,

Thanks for sharing the code! I ran into an issue when trying to reconstruct the image using the released pre-trained weights(vit_base/IN_1M.pth and SOUP_1M_IN.pth). The reconstruction visualization code+hyperparameters follow the original code in MAE: https://github.com/facebookresearch/mae/blob/main/demo/mae_visualize.ipynb

An example of the reconstruction figure(from ImageNet) is attached.

Thanks!

Hi,

Thanks for your great work! I'm interested in reproducing the simulated benchmarks reported in the paper. I've successfully setup the python environment and pretrained vit models. However, it looks like there are no provided training trajectories so that I can reproduce the number in the paper.

Is it possible to release them? Thanks!