https://ankitatiisc.github.io/Strata-NeRF/

https://ankitatiisc.github.io/Strata-NeRF/

Last synced: 3 months ago
JSON representation

• Host: GitHub
• URL: https://ankitatiisc.github.io/Strata-NeRF/
• Owner: ankitatiisc
• License: apache-2.0
• Created: 2023-08-13T16:31:10.000Z (over 1 year ago)
• Default Branch: strata-nerf
• Last Pushed: 2024-02-24T09:51:23.000Z (11 months ago)
• Last Synced: 2024-06-02T01:32:49.081Z (8 months ago)
• Language: Python
• Size: 30.1 MB
• Stars: 10
• Watchers: 1
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-scene-representation - Strata-NeRF: Neural Radiance Fields for Stratified Scenes

README

        
# Strata-NeRF: Neural Radiance Fields for Stratified Scenes

This repository contains the code release for :

[Strata-NeRF](https://ankitatiisc.github.io/Strata-NeRF/),

This implementation is written in [JAX](https://github.com/google/jax), and

is a fork of [mip-NeRF360](https://github.com/google-research/multinerf).

This is research code, and should be treated accordingly.

The link to download our dataset can be found in our [project page](https://ankitatiisc.github.io/Strata-NeRF/) 

## Setup

```

# Clone the repo.

git clone https://github.com/ankitatiisc/Strata-NeRF.git

cd Strata-NeRF

# Make a conda environment.

conda create --name strata_nerf python=3.9

conda activate strata_nerf

# Prepare pip.

conda install pip

pip install --upgrade pip

# Install requirements.

pip install -r requirements.txt

# Manually install rmbrualla's `pycolmap` (don't use pip's! It's different).

git clone https://github.com/rmbrualla/pycolmap.git ./internal/pycolmap

# Confirm that all the unit tests pass.

./scripts/run_all_unit_tests.sh

```

You'll probably also need to update your JAX installation to support GPUs or TPUs.

## Running

Example scripts for training, evaluating, and rendering can be found in

`scripts/`. You'll need to change the paths to point to wherever the datasets

are located. [Gin](https://github.com/google/gin-config) configuration files

for our model and some ablations can be found in `configs/`.

Please set appropriate near and far place values in the config file depending on the number of levels in the scene.

### OOM errors

You may need to reduce the batch size (`Config.batch_size`) to avoid out of memory

errors. If you do this, but want to preserve quality, be sure to increase the number

of training iterations and decrease the learning rate by whatever scale factor you

decrease batch size by.