nomewang / M3DM

MIT License
146 stars 19 forks source link

Different preprocessing in the paper and the code #14

Closed LeeeeTX closed 1 year ago

LeeeeTX commented 1 year ago

When preprocessing, according to your paper, you set the background of point cloud and rgb image to 0. But when I go through the code, I found that in processing.py, you didn't saved the processed rgb image (you commented out these lines). May I ask whether this is intentional or not, or should I run the 124-line "Image.fromarray(organized_clustered_rgb).save(rgb_path)" when I reproduce the experiment? Thanks a lot!

LiZhi-CASIA commented 1 year ago

Hello, I have the same problem as you, did you solve it?

nomewang commented 1 year ago

Thanks for the feedback. This was a bug in the release version and we have removed this comment in the updated code.