Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Official code release for ICCV 2021 paper SNARF: Differentiable Forward Skinning for Animating Non-rigid Neural Implicit Shapes.
| Date | Stars |
|---|---|
| 2026-07-24 | 297 |
| 2026-07-25 | 297 |
| 2026-07-28 | 297 |
| 2026-07-30 | 297 |
| 2026-08-06 | 297 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# SNARF: Differentiable Forward Skinning for Animating Non-rigid Neural Implicit Shapes
## [Paper](https://arxiv.org/pdf/2104.03953.pdf) | [Supp](https://bit.ly/3t1Tk6F) | [Video](https://youtu.be/rCEpFTKjFHE) | [Project Page](https://xuchen-ethz.github.io/snarf) | Blog ([AIT](https://eth-ait.medium.com/animate-implicit-shapes-with-forward-skinning-c7ebbf355694)，[AVG](https://autonomousvision.github.io/snarf/))
<img src="assets/1.gif" width="250" height="250"/> <img src="assets/2.gif" width="250" height="250"/> <img src="assets/3.gif" width="250" height="250"/>
Official code release for ICCV 2021 paper [*SNARF: Differentiable Forward Skinning for Animating Non-rigid Neural Implicit Shapes*](https://arxiv.org/pdf/2104.03953.pdf). We propose a novel forward skinning module to animate neural implicit shapes with good generalization to unseen poses.
**Update:** we have released an improved version, FastSNARF, which is 150x faster than SNARF. Check it out [here](https://github.com/xuchen-ethz/fast-snarf).
If you find our code or paper useful, please cite as
```
@inproceedings{chen2021snarf,
title={SNARF: Differentiable Forward Skinning for Animating Non-Rigid Neural Implicit Shapes},
author={Chen, Xu and Zheng, Yufeng and Black, Michael J and Hilliges, Otmar and Geiger, Andreas},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
```
# Quick Start
Clone this repo:
```
git clone https://github.com/xuchen-ethz/snarf.git
cd snarf
```
Install environment:
```
conda env create -f environment.yml
conda activate snarf
python setup.py install
```
Download [SMPL models](https://smpl.is.tue.mpg.de) (1.0.0 for Python 2.7 (10 shape PCs)) and move them to the corresponding places:
```
mkdir lib/smpl/smpl_model/
mv /path/to/smpl/models/basicModel_f_lbs_10_207_0_v1.0.0.pkl lib/smpl/smpl_model/SMPL_FEMALE.pkl
mv /path/to/smpl/models/basicmodel_m_lbs_10_207_0_v1.0.0.pkl lib/smpl/smpl_model/SMPL_MALE.pkl
```
Download our pretrained models and test motion sequences:
```
sh ./download_data.sh
```
Run a quick demo for clothed human:
```
python demo.py expname=cape subject=3375 demo.motion_path=data/aist_demo/seqs +experiments=cape
```
You can the find the video in `outputs/cape/3375/demo.mp4` and images in `outputs/cape/3375/images/`. To save the meshes, add `demo.save_mesh=true` to the command.
You can also try other subjects (see `outputs/data/cape` for available options) by setting `subject=xx`, and other motion sequences from [AMASS](https://amass.is.tue.mpg.de/download.php) by setting `demo.motion_path=/path/to/amass_modetion.npz`.
Some motion sequences have high fps and one might want to skip some frames. To do this, add `demo.every_n_frames=x` to consider every x frame in the motion sequence. (e.g. `demo.every_n_frames=10` for PosePrior sequences)
By default, we use `demo.fast_mode=true` for fast mesh extraction. In this mode, we first extract mesh in canonical space, and then forward skin the mesh to posed space. This bypasses the root finding during inference, thus is faster. However, it's not really deforming a continuous field. To first deform the continuous field and then extract mesh in deformed space, use `demo.fast_mode=false` instead.
# Training and Evaluation
## Install Additional Dependencies
Install [kaolin](https://kaolin.readthedocs.io/en/latest/notes/installation.html) for fast occupancy query from meshes.
```
git clone https://github.com/NVIDIAGameWorks/kaolin
cd kaolin
git checkout v0.9.0
python setup.py develop
```
## Minimally Clothed Human
### Prepare Datasets
Download the [AMASS](https://amass.is.tue.mpg.de/download.php) dataset. We use ''DFaust Snythetic'' and ''PosePrior'' subsets and SMPL-H format. Unzip the dataset into `data` folder.
```
tar -xf DFaust67.tar.bz2 -C data
tar -xf MPILimits.tar.bz2 -C data
```
Preprocess dataset:
```
python preprocess/sample_points.py --output_folder data/DFaust_processed
python preprocess/sample_points.py --output_folder data/MPI_proceExcerpt of 6,082 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:20f1b1f224d08343, topic:computer-vision, readme:computer vision
matched fp:20f1b1f224d08343, topic:neural-network
matched fp:20f1b1f224d08343, topic:animation