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.
Multi-View Mesh Reconstruction with Neural Deferred Shading (CVPR 2022)
| Date | Stars |
|---|---|
| 2026-07-24 | 268 |
| 2026-07-25 | 269 |
| 2026-07-28 | 269 |
| 2026-07-30 | 269 |
| 2026-08-06 | 269 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<p align="center">
<h1 align="center"><a href="https://fraunhoferhhi.github.io/neural-deferred-shading">Multi-View Mesh Reconstruction with Neural Deferred Shading</a></h1>
<div align="center">
<a href="https://fraunhoferhhi.github.io/neural-deferred-shading">
<img src="docs/static/images/collection_large_bright_small.jpg" alt="Logo" width="60%">
</a>
</div>
<p align="center">
<i>IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022</i>
<br />
<a href="https://mworchel.github.io/"><strong>Markus Worchel</strong></a>*
·
<a href="https://github.com/rodrigodzf"><strong>Rodrigo Diaz</strong></a>*
·
<a href="https://www.hhi.fraunhofer.de/en/vit-imc/team/weiwen-hu.html"><strong>Weiwen Hu</strong></a>
·
Oliver Schreer
·
Ingo Feldmann
·
Peter Eisert
</p>
<p align="center">
*Equal contribution
</p>
</p>
Official code for the CVPR 2022 paper "[Multi-View Mesh Reconstruction with <b>N</b>eural <b>D</b>eferred <b>S</b>hading](https://openaccess.thecvf.com/content/CVPR2022/html/Worchel_Multi-View_Mesh_Reconstruction_With_Neural_Deferred_Shading_CVPR_2022_paper.html)", a method for fast multi-view reconstruction based on differentiable mesh rendering and neural shading.
## Installation
Create a virtual environment and install the dependencies using pip:
```bash
pip install -r requirements.txt
```
### Alternative: Conda
If you are working with conda, you can directly create an environment named `nds` with all dependencies:
```bash
conda env create -f environment.yml
conda activate nds
```
<!-- ### Nvdiffrast
To install [Nvdiffrast](https://github.com/NVlabs/nvdiffrast) from source, run the following in the main directory:
```bash
git clone https://github.com/NVlabs/nvdiffrast.git
cd nvdiffrast
python -m pip install .
``` -->
## Reconstructing DTU Scans
Download the [full dataset](https://www.dropbox.com/s/56ym2qrjfg7jymo/data.zip) (2.3 GB) or [two samples](https://www.dropbox.com/s/x5hrx26l1pmz1id/data.zip) (300 MB) and unzip the content into the main directory. For example, after unzipping you should have the directory `./data/65_skull`.
To start the reconstruction for the skull, run:
```bash
python reconstruct.py --input_dir ./data/65_skull/views --input_bbox ./data/65_skull/bbox.txt
```
or for a general scan:
```bash
python reconstruct.py --input_dir ./data/{SCAN-ID}_{SCAN-NAME}/views --input_bbox ./data/{SCAN-ID}_{SCAN-NAME}/bbox.txt
```
You will find the output meshes in the directory `./out/{SCAN-ID}_{SCAN-NAME}/meshes`.
### Data Conversion from IDR Format to NDS Format
The DTU dataset in the NDS format is derived from the dataset in IDR format ([found here](https://github.com/lioryariv/idr)), which includes masks for a selection of objects. After downloading the dataset from IDR, you can convert it from the IDR format to the NDS format by calling the import script as:
```bash
import_dtu_from_idr.py PATH/TO/IDR/DATASET/DIRECTORY PATH/TO/OUTPUT/DIRECTORY
```
## Reconstructing Custom Scenes
Our pipeline expects the input data in a specific structure, which you have to follow for your own scenes.
### Views (--input_dir)
The main input is a folder with *views*, where each view consists of an RGB(A) image and the corresponding camera pose and camera intrinsics. An example folder with N views could look like this (the views do not have to be numbered and can have any file names):
```bash
📂views
├─🖼️1.png
├─📜1_k.txt
├─📜1_r.txt
├─📜1_t.txt
⋮
├─🖼️N.png
├─📜N_k.txt
├─📜N_r.txt
└─📜N_t.txt
```
If present, the alpha channel of the image is used as object mask.
The files `..._k.txt`, `..._r.txt`, and `..._t.txt` contain numpy-readable arrays with the camera pose (R, t) and intrinsics (K) in the **standard OpenCV format**, so K and R are 3x3 matrices and t is a 3-dimensional column vector, such that
$$ \begin{pmatrix} x & y & 1 \end{pmatrix}^\top \sim \mathbf{K}(\mathbf{R}\begin{pmatrix} X & Y & Z \end{pmatrix}^\top + \mathbf{tExcerpt of 7,931 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:7655dd661ad27599, topic:computer-vision, topic:3d-reconstruction, readme:computer vision
matched fp:7655dd661ad27599, topic:deep-learning, topic:pytorch