Career Profile

I am Junguang Jiang. I graduated from Tsinghua University under the supervision of Mingsheng Long. Now, I work as a machine learning engineer in the display advertising team at Alimama.

I created a Transfer Learning Library TLlib and got over GitHub stars.

Here is my Chinese resume (中文简历) and English resume.


ForkMerge: Overcoming Negative Transfer in Multi-Task Learning
Junguang Jiang, Baixu Chen, Junwei Pan, Ximei Wang, Liu Dapeng, Jie Jiang, Mingsheng Long
Conference and Workshop on Neural Information Processing Systems (NIPS), 2023
Transferability in Deep Learning: A Survey
Junguang Jiang, Yang Shu, Jianmin Wang, Mingsheng Long
arixv, 2022
Debiased Pseudo Labeling in Self-Training
Baixu Chen, Junguang Jiang (co-first authors), Ximei Wang, Jianmin Wang, Mingsheng Long
Conference and Workshop on Neural Information Processing Systems (NIPS Oral), 2022
Decoupled Adaptation for Cross-Domain Object Detection
Junguang Jiang, Baixu Chen, Jianmin Wang, Mingsheng Long
The Tenth International Conference on Learning Representations (ICLR), 2022
Regressive Domain Adaptation for Unsupervised Keypoint Detection
Junguang Jiang, Yifei Ji, Ximei Wang, Yufeng Liu, Jianmin Wang, Mingsheng Long
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2021
Resource Efficient Domain Adaptation
Junguang Jiang, Ximei Wang, Mingsheng Long, and Jianmin Wang
ACM Annual Conference on Multimedia (ACMMM), 2020 (Oral)


Open Source Libraries -- TLlib GitHub stars
  • An open-source and well-documented library for Transfer Learning, Domain Adaptation, Task Adaptation, and Domain Generalization.
  • First Author and contribute more than 20k lines of code.
Selected Course Projects -- TinyCCompiler
  • A tiny C compiler based on LLVM and ANTLR4.
  • Project for Compiler Principles which is scored A+.
  • First Author and responsible for implementing symbol tables, error handling, functions, arrays, pointers, etc.



Tencent Inc, China
  • Research on multi-task learning, mainly for recommendation tasks

Teaching Assistant

Tsinghua University

Teaching Assistant for Machine Learning, Fall 2021


Y-tech, Kuaishou Technology
  • Research on transfer learning for both 3D and 2D keypoint detection, which is then applied to the Augmented Reality (AR) applications.
  • Propose regressive domain adaptation methods for unsupervised keypoint detection to decrease the labeling cost for 3D and 2D keypoints of hand gestures.
  • Related work is published at CVPR2020.

Education & Awards

Master's Degree in Software Engineering

2020 - 2023
Tsinghua University (Beijing, China)
  • Advised by Mingsheng Long
  • Comprehensive Excellence Award (Huawei Scholarship, top 5%), Tsinghua University, 2021

Bachelor's Degree in Software Engineering

2015 - 2020
Tsinghua University (Beijing, China)
  • GPA 3.9 Ranking 2nd/83
  • National Scholarship (top 1%), Tsinghua University, 2019
  • Comprehensive Excellence Award (Huawei Scholarship, top 1%), Tsinghua University, 2018