Created by: IvyGongoogle

Hello, I found one question. When I train model, some warnings show: Found an invalid bbox in annotations: im_id: 89, area: -8902.439024390242 x1: 470.390243902439, y1: 826.0487804878048, x2: 470.390243902439, y2: 914.0731707317073. Obviously, x1=x2. So I track this image's anno json file got by labelme, but find that the all bboxs in this json file is normal, x1≠x2. Like:

{
  "version": "4.5.6",
  "flags": {},
  "shapes": [
    {
      "label": "dog",
      "points": [
        [
          470.390243902439,
          826.0487804878048
        ],
        [
          370.390243902439,
          915.0731707317073
        ]
      ],
      "group_id": null,
      "shape_type": "rectangle",
      "flags": {}
    },
    {
      "label": "dog",
      "points": [
        [
          509.4146341463414,
          858.9756097560975
        ],
        [
          432.5853658536585,
          794.3414634146341
        ]
      ],
      "group_id": null,
      "shape_type": "rectangle",
      "flags": {}
    },
    {
      "label": "dog",
      "points": [
        [
          435.0243902439024,
          751.6585365853658
        ],
        [
          485.0243902439024,
          800.4390243902438
        ]
      ],
      "group_id": null,
      "shape_type": "rectangle",
      "flags": {}
    }
  ],
  "imagePath":xxx
  "imageData": xxx
  "imageHeight": 1080,
  "imageWidth": 1920
}

So what causes this problem.

Can you give some advises? Thanks