RF-DETR Model

RF-DETR is a object detection, segmentation model, and keypoint detection. Built on top of DinoV2 vision transformer backbone. RF-DETR can be trained by two dataset format: COCO and YOLO.

COCO format

dataset/
├── train/
│   ├── _annotations.coco.json
│   ├── image1.jpg
│   ├── image2.jpg
│   └── ... (other image files)
├── valid/
│   ├── _annotations.coco.json
│   ├── image1.jpg
│   ├── image2.jpg
│   └── ... (other image files)
└── test/
    ├── _annotations.coco.json
    ├── image1.jpg
    ├── image2.jpg
    └── ... (other image files)

Structure:

{
  "info": {
    "description": "Dataset description",
    "version": "1.0"
  },
  "licenses": [],
  "images": [
    {
      "id": 1,
      "file_name": "image1.jpg",
      "width": 640,
      "height": 480
    }
  ],
  "categories": [
    {
      "id": 1,
      "name": "cat",
      "supercategory": "animal"
    },
    {
      "id": 2,
      "name": "dog",
      "supercategory": "animal"
    }
  ],
  "annotations": [
    {
      "id": 1,
      "image_id": 1,
      "category_id": 1,
      "bbox": [100, 150, 200, 180],
      "area": 36000,
      "iscrowd": 0
    }
  ]
}