• Per request, we provide two small HRNet models. #parameters and GFLOPs are similar to ResNet18. The segmentation resutls using the two small models are also available at https://github.com/HRNet/HRNet-Semantic-Segmentation.

• TensoFlow implemenation available at https://github.com/yuanyuanli85/tf-hrnet. Thanks VictorLi!

• ONNX export enabled after fixing issues. Thanks Baowen Bao!

This is the official code of high-resolution representations for ImageNet classification. We augment the HRNet with a classification head shown in the figure below. First, the four-resolution feature maps are fed into a bottleneck and the number of output channels are increased to 128, 256, 512, and 1024, respectively. Then, we downsample the high-resolution representations by a 2-strided 3x3 convolution outputting 256 channels and add them to the representations of the second-high-resolution representations. This process is repeated two times to get 1024 channels over the small resolution. Last, we transform 1024 channels to 2048 channels through a 1x1 convolution, followed by a global average pooling operation. The output 2048-dimensional representation is fed into the classifier.

HRNetV2 ImageNet pretrained models are now available!

1. Install PyTorch=0.4.1 following the official instructions
2. git clone https://github.com/HRNet/HRNet-Image-Classification
3. Install dependencies: pip install -r requirements.txt

You can follow the Pytorch implementation: https://github.com/pytorch/examples/tree/master/imagenet

The data should be under ./data/imagenet/images/.

Please specify the configuration file.

For example, train the HRNet-W18 on ImageNet with a batch size of 128 on 4 GPUs:

python tools/train.py --cfg experiments/cls_hrnet_w18_sgd_lr5e-2_wd1e-4_bs32_x100.yaml

For example, test the HRNet-W18 on ImageNet on 4 GPUs:

python tools/valid.py --cfg experiments/cls_hrnet_w18_sgd_lr5e-2_wd1e-4_bs32_x100.yaml --testModel hrnetv2_w18_imagenet_pretrained.pth

• Human pose estimation
• Semantic segmentation
• Object detection
• Facial landmark detection

If you find this work or code is helpful in your research, please cite:

@inproceedings{SunXLW19,
  title={Deep High-Resolution Representation Learning for Human Pose Estimation},
  author={Ke Sun and Bin Xiao and Dong Liu and Jingdong Wang},
  booktitle={CVPR},
  year={2019}
}

@article{WangSCJDZLMTWLX19,
  title={Deep High-Resolution Representation Learning for Visual Recognition},
  author={Jingdong Wang and Ke Sun and Tianheng Cheng and 
          Borui Jiang and Chaorui Deng and Yang Zhao and Dong Liu and Yadong Mu and 
          Mingkui Tan and Xinggang Wang and Wenyu Liu and Bin Xiao},
  journal   = {CoRR},
  volume    = {abs/1908.07919},
  year={2019}
}

[1] Deep High-Resolution Representation Learning for Human Pose Estimation. Ke Sun, Bin Xiao, Dong Liu, and Jingdong Wang. CVPR 2019. download