cdnv2.md

CondenseNet V2

• Paper：CondenseNet V2: Sparse Feature Reactivation for Deep Networks

• Origin Repo：jianghaojun/CondenseNetV2

• Code：cdnv2.py

• Evaluate Transforms：

  # backend: pil
  # input_size: 224x224
  # models: cdnv2_a and cdnv2_b
  transforms = T.Compose([
      T.Resize(256, interpolation='bicubic'),
      T.CenterCrop(224),
      T.ToTensor(),
      T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
  ])
  
  # backend: pil
  # input_size: 224x224
  # models: cdnv2_c
  transforms = T.Compose([
      T.Resize(256, interpolation='bilinear'),
      T.CenterCrop(224),
      T.ToTensor(),
      T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
  ])

• Model Details：

  Model	Model Name	Params (M)	FLOPs (G)	Top-1 (%)	Top-5 (%)	Pretrained Model
  CondenseNetV2-A	cdnv2_a	2.0	0.05	64.38	85.24	Download
  CondenseNetV2-B	cdnv2_b	3.6	0.15	71.89	90.27	Download
  CondenseNetV2-C	cdnv2_c	6.1	0.31	75.87	92.64	Download

• Citation：

  @misc{yang2021condensenet,
      title={CondenseNet V2: Sparse Feature Reactivation for Deep Networks}, 
      author={Le Yang and Haojun Jiang and Ruojin Cai and Yulin Wang and Shiji Song and Gao Huang and Qi Tian},
      year={2021},
      eprint={2104.04382},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
  }