WIP: Neural Network Inductive Biases for High-fidelity 3D Reconstruction

Installation

This project uses uv to manage Python environment and dependencies.

brew install uv   # macOS
# or
pip install uv    # linux, macOS, ...
# or
curl -LsSf https://astral.sh/uv/install.sh | sh

uv will automatically install all dependencies after the first uv run ... command.

Training

# uv run training [HYDRA-OVERRIDES, scene is required]
uv run training scene=thai_statue run_name=first_try

Exporting INR to mesh

# uv run decoding --model-path=[MODEL_PATH]
uv run decoding  --model-path=model_epoch_9000.pt --device=cpu --resolution=64

Resampling pointcloud from the OBJ mesh

# uv run resampling --input-path=[INPUT_OBJ] --num-samples=[NUM_SAMPLES]
uv run resampling  --input-path=<normalized_model.obj --num-samples=1000000

Hydra config's layout

ib/conf
├── config.yaml
├── model
│   ├── inr
│   │   └── siren.yaml
│   ├── loss
│   │   └── siren_sdf_loss.yaml
│   └── original_siren.yaml
├── scene
│   ├── bedroom_1.yaml
│   ├── dataset
│   │   ├── obj_dataset.yaml
│   │   └── xyz_dataset.yaml
│   ├── lamp.yaml
│   ├── siren_room.yaml
│   ├── sofa.yaml
│   ├── sphere.yaml
│   └── thai_statue.yaml
└── trainer
    └── base_trainer.yaml