Instructions to use UEmmanuel5/ProFSAM-Fire-Detector with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ultralytics
How to use UEmmanuel5/ProFSAM-Fire-Detector with ultralytics:
from huggingface_hub import hf_hub_download from ultralytics import YOLO # pick the weights file from this repo's "Files and versions" tab weights = hf_hub_download("UEmmanuel5/ProFSAM-Fire-Detector", "<weights>.pt") model = YOLO(weights) source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True) - Notebooks
- Google Colab
- Kaggle
Upload README.md
Browse files
README.md
CHANGED
|
@@ -53,24 +53,6 @@ train_results = model.train(
|
|
| 53 |
)
|
| 54 |
````
|
| 55 |
|
| 56 |
-
**Dataset filter helper (used to verify class counts):**
|
| 57 |
-
|
| 58 |
-
```python
|
| 59 |
-
import os
|
| 60 |
-
from collections import defaultdict
|
| 61 |
-
|
| 62 |
-
def count_images_by_class(images_dir):
|
| 63 |
-
class_counts = defaultdict(int)
|
| 64 |
-
for filename in os.listdir(images_dir):
|
| 65 |
-
if filename.lower().endswith(('.jpg', '.jpeg', '.png')):
|
| 66 |
-
class_name = filename.split("_CV")[0] if "_CV" in filename else "unknown"
|
| 67 |
-
class_counts[class_name] += 1
|
| 68 |
-
total = sum(class_counts.values())
|
| 69 |
-
print(f"Total images: {total}")
|
| 70 |
-
for cls, count in class_counts.items():
|
| 71 |
-
print(f"{cls}: {count} images")
|
| 72 |
-
```
|
| 73 |
-
|
| 74 |
## Detector metrics (FASDD fire-only subset)
|
| 75 |
|
| 76 |
| P | R | mAP@0.5 | mAP@0.5:0.95 |
|
|
|
|
| 53 |
)
|
| 54 |
````
|
| 55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
## Detector metrics (FASDD fire-only subset)
|
| 57 |
|
| 58 |
| P | R | mAP@0.5 | mAP@0.5:0.95 |
|