Update README.md
Browse files
README.md
CHANGED
|
@@ -11,7 +11,7 @@ FishDet-M is the largest unified benchmark dataset for fish detection in diverse
|
|
| 11 |
|
| 12 |
- **π Dataset Scale**: 105,556 images with 296,885 annotated fish instances
|
| 13 |
- **π Diverse Environments**: Marine, brackish, aquarium, and occluded scenes
|
| 14 |
-
- **π― Unified Format**: COCO-style annotations with bounding boxes
|
| 15 |
- **π€ 28 Models Evaluated**: Including YOLOv8-v12, DETR, R-CNN variants, and more
|
| 16 |
- **π§ CLIP-Based Selection**: Automatic model selection using vision-language alignment
|
| 17 |
|
|
|
|
| 11 |
|
| 12 |
- **π Dataset Scale**: 105,556 images with 296,885 annotated fish instances
|
| 13 |
- **π Diverse Environments**: Marine, brackish, aquarium, and occluded scenes
|
| 14 |
+
- **π― Unified Format**: COCO-style annotations with bounding boxes
|
| 15 |
- **π€ 28 Models Evaluated**: Including YOLOv8-v12, DETR, R-CNN variants, and more
|
| 16 |
- **π§ CLIP-Based Selection**: Automatic model selection using vision-language alignment
|
| 17 |
|