Log In Sign Up
Models:
David Feaster
DavidFeaster/
Surgical-Bleeding-AMAGI
0
Downloads: 1
Diff of /configs/detection/acrn/metafile.yml [000000] .. [6d389a]

Switch to unified view

a b/configs/detection/acrn/metafile.yml 

  1
  
Collections:





  
  

  2
  
- Name: ACRN





  
  

  3
  
  README: configs/detection/acrn/README.md





  
  

  4
  
  Paper:





  
  

  5
  
    URL: https://arxiv.org/abs/1807.10982





  
  

  6
  
    Title: Actor-Centric Relation Network





  
  

  7
  
Models:





  
  

  8
  
- Config: configs/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb.py





  
  

  9
  
  In Collection: ACRN





  
  

  10
  
  Metadata:





  
  

  11
  
    Architecture: ResNet50





  
  

  12
  
    Batch Size: 6





  
  

  13
  
    Epochs: 10





  
  

  14
  
    Input: 32x2





  
  

  15
  
    Modality: RGB





  
  

  16
  
    Parameters: 92232057





  
  

  17
  
    Pretrained: Kinetics-400





  
  

  18
  
    Training Data: AVA v2.1





  
  

  19
  
    Training Resources: 8 GPUs





  
  

  20
  
  Name: slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb





  
  

  21
  
  Results:





  
  

  22
  
  - Dataset: AVA v2.1





  
  

  23
  
    Metrics:





  
  

  24
  
      mAP: 27.1





  
  

  25
  
    Task: Spatial Temporal Action Detection





  
  

  26
  
  Training Json Log: https://download.openmmlab.com/mmaction/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb.json





  
  

  27
  
  Training Log: https://download.openmmlab.com/mmaction/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb.log





  
  

  28
  
  Weights: https://download.openmmlab.com/mmaction/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava_rgb-49b07bf2.pth





  
  

  29
  
- Config: configs/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb.py





  
  

  30
  
  In Collection: ACRN





  
  

  31
  
  Metadata:





  
  

  32
  
    Architecture: ResNet50





  
  

  33
  
    Batch Size: 6





  
  

  34
  
    Epochs: 10





  
  

  35
  
    Input: 32x2





  
  

  36
  
    Modality: RGB





  
  

  37
  
    Parameters: 92232057





  
  

  38
  
    Pretrained: Kinetics-400





  
  

  39
  
    Training Data: AVA v2.2





  
  

  40
  
    Training Resources: 8 GPUs





  
  

  41
  
  Name: slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb





  
  

  42
  
  Results:





  
  

  43
  
  - Dataset: AVA v2.2





  
  

  44
  
    Metrics:





  
  

  45
  
      mAP: 27.8





  
  

  46
  
    Task: Spatial Temporal Action Detection





  
  

  47
  
  Training Json Log: https://download.openmmlab.com/mmaction/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb.json





  
  

  48
  
  Training Log: https://download.openmmlab.com/mmaction/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb.log





  
  

  49
  
  Weights: https://download.openmmlab.com/mmaction/detection/acrn/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb/slowfast_acrn_kinetics_pretrained_r50_8x8x1_cosine_10e_ava22_rgb-2be32625.pth