Junkun Jiang (姜俊锟)

---

I am now a senior algorithm engineer working for BIGO Technology, a short video provider(Likee) and a global live video social platform(Bigolive). My job is to develop enhancing algorithms based on image processing and video processing.

I am interested in Computer Vision and Image Processing, especially in sketch-based image retrieval, sketch segmentation, and 3D model segmentation. Recently, I am working on saliency detection and similar object segmentation technologies for video stream bitrate reducing and detail enhancement. Due to the deploy requirement, I also have the ability to accelerate the NN model via TensorRT and OpenVINO, and Neon intrinsic based on mobile deployment tasks. Besides, I have a little bit of understanding of the prune and light-weighted methods.

In my spare time, I enjoy jogging and photography. I love basketball.

Work Experience

• BIGO TECHNOLOGY PTE. LTD. Guangzhou, China
  • Image Algorithm Engineer Jul. 2019 - Present
  • Image Algorithm Engineer Intern Jun. 2018 - Aug. 2018

Education

• M.Eng in Software Engineering, Sun Yat-Sen University, 2019.
  (National Engineering Research Center of Digital Life, Guangzhou, China)
  Co-supervised by Prof. Ruomei Wang and A.P. Shujin Lin
• B.Eng in Computer Science and Engineering, Xiamen University, 2017.
  Thesis advised by A.P. Liansheng Wang

Publications

1. Fei Wang*, Yu Yang, Baoquan Zhao, Junkun Jiang, Teng Zhou, Dazhi Jiang and Tie Cai. Deep 3D Shape Reconstruction from Single-View Sketch Image[C]// 2020 International Conference on Digital Home (ICDH). IEEE, 2020. (preprint)
2. Junkun Jiang, Ruomei Wang*, Shujin Lin and Fei Wang. SFSegNet: Parse Freehand Sketches using Deep Fully Convolutional Networks[C]// 2019 International Joint Conference on Neural Networks (IJCNN). IEEE, 2019: 1-8. (preprint) (DOI)
3. Fei Wang, Shujin Lin*, Hanhui Li, Hefeng Wu, Junkun Jiang, Ruomei Wang, Xiaonan Luo. Multi-column Point-CNN for Sketch Segmentation[J]. Neurocomputing, 2019. (preprint)
4. Ruomei Wang, Junkun Jiang, Shujin Lin* and Fei Wang. An Edge Feature Extraction Algorithm for Sketch Retrieval[C]// The 12th Conference on China Computer Graphics (Chinagraph), 2018. (P000108) (preprint)

Patents

1. Junkun Jiang, Shujin Lin* and Fan Zhou. 2019. An image retrieval method and system based on hand-drawn style recommendation. CN109308324A, filed 2018-09-08, and issued 2019-02-05.
2. Fei Wang, Shujin Lin*, Fan Zhou and Junkun Jiang. 2019. An image retrieval method and system based on hand-drawn contour map matching. CN109145140A, filed 2018-09-08, issued 2019-01-04.

Honor

1. Scientific and Technological Cooperation Award (10%), National Engineering Research Center of Digital Life, Sun Yat-Sen University, 2018-2019
2. Outstanding graduate thesis Award (10%), School of information science and engineering, Xiamen University, 2016-2017

More

3. Third-class National Scholarship **(50%)**, China's Minister of Education, 2018-2019 4. Third-class National Scholarship **(50%)**, China's Minister of Education, 2017-2018 5. Merit Student **(20%)**, Xiamen University, 2016-2017 6. Second-class Scholarship **(20%)**, Xiamen University, 2014-2015 7. Second-class Scholarship **(20%)**, Xiamen University, 2013-2014 8. Cultural and sports activities Scholarship **(20%)**, Xiamen University, 2013-2014 9. Outstanding students of military training **(10%)**, Xiamen University, 2013-2014

Programming

• Languages
  • C/C++
  • Python
  • Matlab
  • Neon intrinsic
• NN tools
  • PyTorch/ONNX
  • Caffe
• NN deploying
  • TensorRT
  • OpenVINO