metafile.yml

Collections:
  - Name: DFF
    Metadata:
      Training Data: ImageNet DET, ImageNet VID
      Training Techniques:
        - SGD with Momentum
      Training Resources: 8x V100 GPUs
      Architecture:
        - ResNet
    Paper:
        URL: https://arxiv.org/abs/1611.07715
        Title: Deep Feature Flow for Video Recognition
    README: configs/vid/dff/README.md

Models:
  - Name: dff_faster_rcnn_r50_dc5_1x_imagenetvid
    In Collection: DFF
    Config: configs/vid/dff/dff_faster_rcnn_r50_dc5_1x_imagenetvid.py
    Metadata:
      Training Memory (GB): 2.50
    Results:
      - Task: Video Object Detection
        Dataset: ImageNet VID
        Metrics:
          box AP@0.5: 70.3
    Weights: https://download.openmmlab.com/mmtracking/vid/dff/dff_faster_rcnn_r50_dc5_1x_imagenetvid/dff_faster_rcnn_r50_dc5_1x_imagenetvid_20201227_213250-548911a4.pth

  - Name: dff_faster_rcnn_r101_dc5_1x_imagenetvid
    In Collection: DFF
    Config: configs/vid/dff/dff_faster_rcnn_r101_dc5_1x_imagenetvid.py
    Metadata:
      Training Memory (GB): 3.25
    Results:
      - Task: Video Object Detection
        Dataset: ImageNet VID
        Metrics:
          box AP@0.5: 73.5
    Weights: https://download.openmmlab.com/mmtracking/vid/dff/dff_faster_rcnn_r101_dc5_1x_imagenetvid/dff_faster_rcnn_r101_dc5_1x_imagenetvid_20201218_172720-ad732e17.pth

  - Name: dff_faster_rcnn_x101_dc5_1x_imagenetvid
    In Collection: DFF
    Config: configs/vid/dff/dff_faster_rcnn_x101_dc5_1x_imagenetvid.py
    Metadata:
      Training Memory (GB): 4.95
    Results:
      - Task: Video Object Detection
        Dataset: ImageNet VID
        Metrics:
          box AP@0.5: 75.5
    Weights: https://download.openmmlab.com/mmtracking/vid/dff/dff_faster_rcnn_x101_dc5_1x_imagenetvid/dff_faster_rcnn_x101_dc5_1x_imagenetvid_20210819_095932-0a9e6cb5.pth