Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ducha-aiki/pytorch-sift
PyTorch implementation of SIFT descriptor
https://github.com/ducha-aiki/pytorch-sift
cnn descriptor image-matching local-features pytorch sift
Last synced: 2 months ago
JSON representation
PyTorch implementation of SIFT descriptor
- Host: GitHub
- URL: https://github.com/ducha-aiki/pytorch-sift
- Owner: ducha-aiki
- License: bsd-2-clause
- Created: 2017-04-30T18:29:05.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2019-08-22T10:44:48.000Z (over 5 years ago)
- Last Synced: 2024-01-28T23:10:57.101Z (11 months ago)
- Topics: cnn, descriptor, image-matching, local-features, pytorch, sift
- Language: Jupyter Notebook
- Size: 1.08 MB
- Stars: 165
- Watchers: 4
- Forks: 16
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
This is an differentiable [pytorch](https://github.com/pytorch/pytorch) implementation of SIFT patch descriptor. It is very slow for describing one patch, but quite fast for batch. It can be used for descriptop-based learning shape of affine feature.
**UPD 08/2019** : pytorch-sift is added to [kornia](https://github.com/arraiyopensource/kornia) and available by `kornia.features.SIFTDescriptor`
There are different implementations of the SIFT on the web. I tried to match [Michal Perdoch implementation](https://github.com/perdoch/hesaff/blob/master/siftdesc.cpp), which gives high quality features for image retrieval [CVPR2009](http://cmp.felk.cvut.cz/~chum/papers/perdoch-cvpr09.pdf). However, on planar datasets, it is inferior to [vlfeat implementation](http://www.vlfeat.org/sandbox/api/sift.html).
The main difference is gaussian weighting window parameters, so I have made a vlfeat-like version too. MP version weights patch center much more (see image below, left) and additionally crops everything outside the circular region. Right is vlfeat version
```python
descriptor_mp_mode = SIFTNet(patch_size = 65,
sigma_type= 'hesamp',
masktype='CircularGauss')
descriptor_vlfeat_mode = SIFTNet(patch_size = 65,
sigma_type= 'vlfeat',
masktype='Gauss')
```
Results:
```
OPENCV-SIFT - mAP
Easy Hard Tough mean
------- ------- --------- -------
0.47788 0.20997 0.0967711 0.26154
VLFeat-SIFT - mAP
Easy Hard Tough mean
-------- -------- --------- --------
0.466584 0.203966 0.0935743 0.254708
PYTORCH-SIFT-VLFEAT-65 - mAP
Easy Hard Tough mean
-------- -------- --------- --------
0.472563 0.202458 0.0910371 0.255353
NUMPY-SIFT-VLFEAT-65 - mAP
Easy Hard Tough mean
-------- -------- --------- --------
0.449431 0.197918 0.0905395 0.245963
PYTORCH-SIFT-MP-65 - mAP
Easy Hard Tough mean
-------- -------- --------- --------
0.430887 0.184834 0.0832707 0.232997
NUMPY-SIFT-MP-65 - mAP
Easy Hard Tough mean
-------- ------- --------- --------
0.417296 0.18114 0.0820582 0.226832
```
Speed:
- 0.00246 s per 65x65 patch - [numpy SIFT](https://github.com/ducha-aiki/numpy-sift)
- 0.00028 s per 65x65 patch - [C++ SIFT](https://github.com/perdoch/hesaff/blob/master/siftdesc.cpp)
- 0.00074 s per 65x65 patch - CPU, 256 patches per batch
- 0.00038 s per 65x65 patch - GPU (GM940, mobile), 256 patches per batch
- 0.00038 s per 65x65 patch - GPU (GM940, mobile), 256 patches per batch
If you use this code for academic purposes, please cite the following paper:
```
@InProceedings{AffNet2018,
title = {Repeatability Is Not Enough: Learning Affine Regions via Discriminability},
author = {Dmytro Mishkin, Filip Radenovic, Jiri Matas},
booktitle = {Proceedings of ECCV},
year = 2018,
month = sep
}
```