search

mabaorui / NeuralPull

Implementation of ICML'2021:Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces
MIT License
179 stars 28 forks source link

Data for Single-view Reconstruction #16

Open Wi-sc opened 1 year ago

Wi-sc commented 1 year ago

Hi @mabaorui, thanks for your code. I'm interested in the single-view reconstruction part and I want to train it on new data. But before that I want to reproduce it.

I noticed that the image features are pre-extracted by a pre-trained model from SoftRas as you mentioned in the issue #4. But they released three models with silhouettes supervision, with shading supervision and with color supervision. Which one was used here? Or could you provide the image feature data here if possible? Thank you again.

Wi-sc commented 1 year ago

Hi @mabaorui, I think your single-view reconstruction under chamfer distance L1 comparison is unfair. Results from Occupancy Networks are scaled 10 but you forget the scale. Please check it here. https://github.com/autonomousvision/occupancy_networks/issues/61#issue-617865443