metafile.yml

Collections:
  - Name: VoteNet
    Metadata:
      Training Techniques:
        - AdamW
      Training Resources: 8x V100 GPUs
      Architecture:
        - PointNet++
    Paper: https://arxiv.org/abs/1904.09664
    README: configs/votenet/README.md

Models:
  - Name: votenet_16x8_sunrgbd-3d-10class.py
    In Collection: VoteNet
    Config: configs/votenet/votenet_16x8_sunrgbd-3d-10class.py
    Metadata:
      Training Data: SUNRGBD
      Training Memory (GB): 8.1
    Results:
      - Task: 3D Object Detection
        Dataset: SUNRGBD
        Metrics:
          AP@0.25: 59.07
          AP@0.5: 35.77
    Weights: https://download.openmmlab.com/mmdetection3d/v0.1.0_models/votenet/votenet_16x8_sunrgbd-3d-10class/votenet_16x8_sunrgbd-3d-10class_20200620_230238-4483c0c0.pth

  - Name: votenet_8x8_scannet-3d-18class.py
    In Collection: VoteNet
    Config: configs/votenet/votenet_8x8_scannet-3d-18class.py
    Metadata:
      Training Data: ScanNet
      Training Memory (GB): 4.1
    Results:
      - Task: 3D Object Detection
        Dataset: ScanNet
        Metrics:
          AP@0.25: 62.90
          AP@0.5: 39.91
    Weights: https://download.openmmlab.com/mmdetection3d/v0.1.0_models/votenet/votenet_8x8_scannet-3d-18class/votenet_8x8_scannet-3d-18class_20200620_230238-2cea9c3a.pth

  - Name: votenet_iouloss_8x8_scannet-3d-18class
    In Collection: VoteNet
    Config: configs/votenet/votenet_iouloss_8x8_scannet-3d-18class.py
    Metadata:
      Training Data: ScanNet
      Training Memory (GB): 4.1
      Architecture:
        - IoU Loss
    Results:
      - Task: 3D Object Detection
        Dataset: ScanNet
        Metrics:
          AP@0.25: 63.81
          AP@0.5: 44.21