|
--- |
|
license: mit |
|
library_name: ultralytics |
|
tags: |
|
- yolov8 |
|
- object-detection |
|
- pytorch |
|
--- |
|
|
|
# TabDetect |
|
|
|
### Supported Labels |
|
|
|
``` |
|
['full_lined', 'not_full_lined'] |
|
``` |
|
|
|
### How to use |
|
|
|
- Install ultralytics: |
|
|
|
```bash |
|
pip install -U ultralytics==8.0.227 |
|
``` |
|
|
|
- Load model and perform prediction: |
|
|
|
```python |
|
from ultralytics import YOLO |
|
|
|
# load model |
|
model = YOLO('camiloa2m/TabDetect-YOLOv8s') |
|
|
|
# set model parameters (optional) |
|
model.overrides['conf'] = 0.25 # NMS confidence threshold |
|
model.overrides['iou'] = 0.45 # NMS IoU threshold |
|
model.overrides['agnostic_nms'] = False # NMS class-agnostic |
|
model.overrides['max_det'] = 1000 # maximum number of detections per image |
|
|
|
# set image |
|
image = '<URL or Path to an image' |
|
|
|
# perform inference |
|
results = model.predict(image) |
|
``` |
|
|
|
### Dataset |
|
|
|
[TNCR_Dataset](https://github.com/abdoelsayed2016/TNCR_Dataset). I merged some classes: class 0 (full_lined, merged_cells), class 1 (no_lines, partial_lined, partial_lined_merged_cells). |
|
|
|
### Model summary (fused) |
|
|
|
| Class | Images | Instances | P | R | mAP50 | mAP50-95 | |
|
|----------------|--------|-----------|-------|-------|-------|----------| |
|
| all | 1313 | 1906 | 0.957 | 0.926 | 0.973 | 0.938 | |
|
| full_lined | 1313 | 984 | 0.96 | 0.949 | 0.98 | 0.968 | |
|
| not_full_lined | 1313 | 922 | 0.953 | 0.904 | 0.966 | 0.908 | |
|
|
|
|
|
<div align="center"> |
|
<img width="640" alt="camiloa2m/TabDetect-YOLOv8n" src="https://huggingface.co/camiloa2m/TabDetect-YOLOv8n/resolve/main/val_batch1_pred.jpg"> |
|
</div> |
|
|
|
|
