High-Resolution Representation Learning for ImageNet Classification

Ke Sun     Yang Zhao     Borui Jiang     Tianheng Cheng     Bin Xiao     Dong Liu     Yadong Mu     Xinggang Wang     Wenyu Liu     Jingdong Wang


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.
The code and models are publicly available at GitHub.





We released the training and testing code and the pretrained model at GitHub

Other applications

more ...
Pose estimation Semantic segmentation Face alignment Image classification Object detection


  title={Deep High-Resolution Representation Learning for Human Pose Estimation},
  author={Ke Sun and Bin Xiao and Dong Liu and Jingdong Wang},
  title={High-Resolution Representations for Labeling Pixels and Regions},
  author={Ke Sun and Yang Zhao and Borui Jiang and Tianheng Cheng and Bin Xiao
  and Dong Liu and Yadong Mu and Xinggang Wang and Wenyu Liu and Jingdong Wang},
  journal   = {CoRR},
  volume    = {abs/1904.04514},


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

HTML Hit Counters