Update README
Browse files
README.md
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- kornia
|
| 5 |
+
- feature-detection
|
| 6 |
+
- feature-description
|
| 7 |
+
- image-matching
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# kornia/disk
|
| 11 |
+
|
| 12 |
+
Pretrained weights for **DISK** (Learning local features with policy gradient),
|
| 13 |
+
used by [`kornia.feature.DISK`](https://kornia.readthedocs.io/en/latest/feature.html).
|
| 14 |
+
|
| 15 |
+
DISK is trained with policy gradient and produces dense heatmaps of keypoint
|
| 16 |
+
probability and per-pixel 128-dim descriptors via a U-Net backbone. NeurIPS 2020.
|
| 17 |
+
|
| 18 |
+
**Original repo:** [cvlab-epfl/disk](https://github.com/cvlab-epfl/disk)
|
| 19 |
+
|
| 20 |
+
## Weights
|
| 21 |
+
|
| 22 |
+
| File | Training supervision |
|
| 23 |
+
|------|----------------------|
|
| 24 |
+
| `depth-save.pth` | Depth |
|
| 25 |
+
| `epipolar-save.pth` | Epipolar geometry |
|
| 26 |
+
|
| 27 |
+
## Citation
|
| 28 |
+
|
| 29 |
+
```bibtex
|
| 30 |
+
@article{tyszkiewicz2020disk,
|
| 31 |
+
title = {{DISK}: Learning local features with policy gradient},
|
| 32 |
+
author = {Tyszkiewicz, Micha{\l} and Fua, Pascal and Trulls, Eduard},
|
| 33 |
+
journal = {Advances in Neural Information Processing Systems},
|
| 34 |
+
volume = {33},
|
| 35 |
+
pages = {14254--14265},
|
| 36 |
+
year = {2020}
|
| 37 |
+
}
|
| 38 |
+
```
|