AttributeError: 'MMDetection' object has no attribute 'value'

[zym3122]

I loaded the SAM back-end in the background, and successfully connected the reasoning model in the front page. However, when marking, the sam reasoning model could not be automatically recognized according to the prompt points or rectangular boxes, and an error was reported in the terminal: ‘’AttributeError: 'MMDetection' object has no attribute 'value'‘

[JimmyMa99]

This error looks like it happens when the LS backend reads based on the image url, ask if the backend inference and the dataset are in the same device.

[zym3122]

解决了，但是具体不知道为什么，我重新拉取安装了一下，重新连接并新建后台项目就可以了。PS:(应该是之前没配置好)

[JimmyMa99]

解决了就好哈哈

[ccly1996]

解决了就好哈哈

我也碰到这个问题了，重装了也不行

[JamesBonddu]

传递过来的task[‘data']格式如下

{'id': 210, 'data': {'image': '/data/upload/3/f43a4c65-IMG_20210627_225110.jpg'}, 'meta': {}, 'created_at': '2023-09-05T02:43:47.050202Z', 'updated_at': '2023-09-05T06:46:57.134849Z', 'is_labeled': True, 'overlap': 1, 'inner_id': 1, 'total_annotations': 1, 'cancelled_annotations': 0, 'total_predictions': 0, 'comment_count': 0, 'unresolved_comment_count': 0, 'last_comment_updated_at': None, 'project': 3, 'updated_by': 1, 'file_upload': 211, 'comment_authors': [], 'annotations': [{'id': 4, 'created_username': ' [email protected], 1', 'created_ago': '20\xa0minutes', 'completed_by': 1, 'result': [{'original_width': 4000, 'original_height': 3000, 'image_rotation': 0, 'value': {'x': 28.655660377358487, 'y': 38.05031446540881, 'width': 35.25943396226414, 'height': 33.9622641509434, 'rotation': 0, 'rectanglelabels': ['cat']}, 'id': 'ekRgHmHoPQ', 'from_name': 'RectangleLabels', 'to_name': 'image', 'type': 'rectanglelabels', 'origin': 'manual'}], 'was_cancelled': False, 'ground_truth': False, 'created_at': '2023-09-05T06:46:30.854028Z', 'updated_at': '2023-09-05T06:46:57.089316Z', 'draft_created_at': '2023-09-05T03:10:53.669362Z', 'lead_time': 26.183, 'import_id': None, 'last_action': None, 'task': 210, 'project': 3, 'updated_by': 1, 'parent_prediction': None, 'parent_annotation': None, 'last_created_by': None}], 'predictions': []} kwargs {'login': None, 'password': None, 'context': None}

所以我把self.value调整为 .get("image")