Skip to content

Latest commit

 

History

History
21 lines (11 loc) · 943 Bytes

README.md

File metadata and controls

21 lines (11 loc) · 943 Bytes

Training Template

A template to perform the PyTorch training, maybe saving some time.

TODO

  • Implement Multi-GPU version. The current version only support single GPU training. For multiple GPUs, refer to NPVA/train.py.

Usage

  1. The training program begins from train.py.

  2. You just follow the running stream and fill each 'TODO' thing.

Methods using This Template

  1. [link] MeGA: Hybrid Mesh-Gaussian Head Avatar for High-Fidelity Rendering and Head Editing, arXiv'24.

  2. [link] Neural Point-based Volumetric Avatar: Surface-guided Neural Points for Efficient and Photorealistic Volumetric Head Avatar, SIGGRAPH Asia 2023.

  3. [link] MotionHint: Self-Supervised Monocular Visual Odometry with Motion Constraints, ICRA'22.