Update README.md
Browse files
README.md
CHANGED
|
@@ -24,9 +24,11 @@ Trained Model based on [yolov9-e](https://github.com/WongKinYiu/yolov9) is in [h
|
|
| 24 |
|
| 25 |
|
| 26 |
# Statistics
|
| 27 |
-
|
| 28 |
- Train: 3,991 images
|
| 29 |
- Val: 399 images
|
|
|
|
|
|
|
|
|
|
| 30 |
|
| 31 |
| Character Name | # in Train | # in Val | Precision | Recall | mAP50 | mAP50-95 |
|
| 32 |
|:------------------:|:---:|:---:|:---:|:---:|:---:|:---:|
|
|
|
|
| 24 |
|
| 25 |
|
| 26 |
# Statistics
|
|
|
|
| 27 |
- Train: 3,991 images
|
| 28 |
- Val: 399 images
|
| 29 |
+
- All labels have been manually created through human effort, with workers ethically compensated with "passion pay", came from south korea.
|
| 30 |
+
- Each image is named to indicate its source.
|
| 31 |
+
|
| 32 |
|
| 33 |
| Character Name | # in Train | # in Val | Precision | Recall | mAP50 | mAP50-95 |
|
| 34 |
|:------------------:|:---:|:---:|:---:|:---:|:---:|:---:|
|