Yuhang He's Blog

Some birds are not meant to be caged, their feathers are just too bright.

Conference Publications

Link to [Google Scholar]

Real-Time Fashion-guided Clothing Semantic Parsing: a Lightweight Multi-Scale Inception Neural Network and Benchmark
Yuhang He, Lu Yang, Long Chen. [PDF]
The Thirty-First AAAI Conference on Artificial Intelligence (AAAI-17 workshop), 2017.

Fast Fashion Guided Clothing Image Retrieval: Delving Deeper into What Feature Makes Fashion
Yuhang He, Long Chen.[PDF]
The 13th Asian Conference on Computer Vision (ACCV16), 2016.
We test various tradtional hand-crafed and deep metric learning feature to figure out what feature makes fashion.

Multi-task Relative Attribute Prediction by Incorporating Local Context and Global Style Information
Yuhang He, Long Chen, Jianda Chen. (Oral) [PDF]
The 27th British Machine Vision Conference (BMVC 16), 2016.
We propose an end-to-end trainable deep neural network to predict the relative attribute between an image pair by leveraging both local feature and global style information.

A Novel Way to Organize 3D LiDAR Point Cloud as 2D Depth Map Height Map and Surface Normal Map
Yuhang He, Long Chen, Ming Li. [PDF] [Code]
IEEE International Conference on Robotics and Biomimetics (ROBIO15), 2015.
We propose a parameter self-adaptive framework to organize 3D LiDAR point cloud as 2D dense depth map, height map and surface normal map under the guidance of RGB image.

Robust Optimization with Credibility Factor for Graph-based SLAM
Long Chen, Yuhang He, Kai Huang. [PDF]
IEEE International Conference on Robotics and Biomimetics (ROBIO15), 2015.
We propose a parameter self-adaptive framework to organize 3D LiDAR point cloud as 2D dense depth map, height map and surface normal map under the guidance of RGB image.

Sparse Depth Map Upsampling with RGB Image and Anisotropic Diffusion Tensor
Yuhang He, Long Chen, Ming Li. [PDF] [Code]
IEEE 2015 Intelligent Vehicle Symposium (IV 15).
3D LiDAR point cloud projects to image plane and produces sparse depth map. To upgrade its resolution, we provide an upsampling framework by integrating RGB image and anisotropic diffusion tensor.

Using Edit Distance and Junction Feature to Dedect and Recognize Arrow Road Marking
Yuhang He, Shi Chen, Yifeng Pan, Kai Ni. [PDF] [Code]
IEEE 2014 Intelnational Conference on Intelligent Transportation Systems (ITSC 14), 2014.
To detect and classify arrow road marking, we provide to express each arrow road as junction feature string and its similarity with marking templates is measured by edit distance.

Journal Publications

Transforming 3D LiDAR Point Cloud into 2D Dense Depth Map through a Parameter Self-adaptive Framework
Long Chen, Yuhang He. (corresponding author) [PDF]
IEEE Transactions on Intelligent Transportation Systems (TITS), 2016
We design a parameter self-adaptive framework to organize 3D LiDAR point cloud into 2D dense depth map.