Skip to content

Explored view consistant portrait image generation and semantic manipulation using CNeRF

Notifications You must be signed in to change notification settings

HemaDevaSagar35/CNeRF

Repository files navigation

CNeRF

model can be found here create a folder called checkpoints and put the model in there Below are the instruction to run the application

  1. Make sure you have pytorch and pytorch 3D. For Pytorch 3D you have to do source install. You can find relevant detail here.Note Installing this is a very long process. Like compiling the source code takes close to 15 - 20 minutes. Please be patient.
  2. Do pip install -r requirement.txt
  3. Go to the app folder and run the application as
    python app.py

If you want to do training from scratch, you have to run the command
python train_CNeRF_modified.py

But before running this command, you have to adjust the dataset directory parameter dataset_path and checkpoints_dir parameter in options.py

Also you can find the dataset I used here

Project link: For more details check the project details webpage.

References
[1] https://github.com/MrTornado24/FENeRF
[2] https://github.com/TianxiangMa/Compositional-NeRF
[3] https://arxiv.org/pdf/2302.01579.pdf

About

Explored view consistant portrait image generation and semantic manipulation using CNeRF

Topics

Resources

Stars

Watchers

Forks