apple / ml-neuman

Official repository of NeuMan: Neural Human Radiance Field from a Single Video (ECCV 2022)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

The problem with render_reposing.py

xiaochenwise opened this issue · comments

I use your trained model to reason my own data with the following command: 

python render_reposing.py --scene_dir data/mydata/output --weights_path out/lab_human/checkpoint.pth.tar --motion_name=jumpandroll.
the result for a frame is as follows:
out_0007
The people in the picture above are still the people in the training set, not the people in my data set. What is the problem?

You need to retrain the scene model and human model for your custom data, you've only preprocessed the data, not trained any model from it.