Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,4 +1,35 @@
|
|
| 1 |
---
|
| 2 |
-
library_name: image-matching-models
|
| 3 |
license: apache-2.0
|
| 4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- image-matching
|
| 5 |
+
- computer-vision
|
| 6 |
+
- feature-matching
|
| 7 |
+
library_name: image-matching-models
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# XoFTR
|
| 11 |
+
|
| 12 |
+
XoFTR is a cross-view feature matching transformer for image matching.
|
| 13 |
+
|
| 14 |
+
## Files
|
| 15 |
+
|
| 16 |
+
- `xoftr_640.safetensors`: XoFTR model pretrained at 640px resolution
|
| 17 |
+
- `xoftr_840.safetensors`: XoFTR model pretrained at 840px resolution
|
| 18 |
+
|
| 19 |
+
## Usage
|
| 20 |
+
|
| 21 |
+
```python
|
| 22 |
+
from matching import get_matcher
|
| 23 |
+
|
| 24 |
+
# Use 640 variant (default)
|
| 25 |
+
matcher = get_matcher("xoftr", device="cuda", pretrained_size=640)
|
| 26 |
+
|
| 27 |
+
# Or use 840 variant
|
| 28 |
+
matcher = get_matcher("xoftr", device="cuda", pretrained_size=840)
|
| 29 |
+
|
| 30 |
+
mkpts0, mkpts1, _, _, _, _ = matcher(img0, img1)
|
| 31 |
+
```
|
| 32 |
+
|
| 33 |
+
## Links
|
| 34 |
+
- [Original repository](https://github.com/OnderT/XoFTR)
|
| 35 |
+
- [Image Matching Models](https://github.com/gmberton/image-matching-models)
|