|
|
@@ -216,8 +216,8 @@ class VOCDetection(Dataset):
|
|
|
annotations['annotations'].append({
|
|
|
'iscrowd': 0,
|
|
|
'image_id': int(im_id[0]),
|
|
|
- 'bbox': [x1, y1, x2 - x1 + 1, y2 - y1 + 1],
|
|
|
- 'area': float((x2 - x1 + 1) * (y2 - y1 + 1)),
|
|
|
+ 'bbox': [x1, y1, x2 - x1, y2 - y1],
|
|
|
+ 'area': float((x2 - x1) * (y2 - y1)),
|
|
|
'category_id': cname2cid[cname] + 1,
|
|
|
'id': ann_ct,
|
|
|
'difficult': _difficult
|