I am leading the Ecosystem and User Experience algorithm team at TikTok Business Integrity, where we build strong capabilities to detect risks and user experience issues on TikTok’s monetization ecosystem. We strive for making TikTok safe to gain user trust, enabling sustainable business growth.

Prior to TikTok, I was a Research Scientist at Gaussian Robotics, where worked on various topics on intelligent robotics, such as visual SLAM related deep learning algoirthms, multi-task learning and efficient deep learning model inference on edge-devices. I also spent time at ViSenze and Tuputech, leading various computer vision projects.

I obtained my bachelor’s degree in Computing Science from the University of Glasgow in 2016, after spending 3 wonderful years in Scotland, I also obtained a bachelor’s degree in Eletronics Engineering from Sun Yat-Sen University in the same year.

I have broad interests in AI. Currently, I am working on building strong LLMs and MLLMs to empower TikTok’s moderation system via both pre-training and post-training. I am also interested in 3D computer vision problems and building systems/algorithms for autonomous robotics that efficiently interact with our physical world.

[Hiring] I am actively hiring talented research scientists/engineers in Singapore, San Jose and mainland of China to work on the next generation moderation system. Please feel free to reach out if you are interested in working on impactful projects and challenging problems!

News

• [07.2022] Our work “A Deep-Learning-based System for Indoor Active Cleaning” has been accepted to IROS 2022.
• [07.2021] Our work “MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesis” has been accepted to ICCV 2021.
• [10.2019] I presented our winning solution to the DeepFashion2 cloth retrieval challenge at ICCV in Seoul.

Publications

• A Deep-Learning-based System for Indoor Active Cleaning
  Yike Yun^*, Linjie Hou^*, Zijian Feng^*, Wei Jin, Yang Liu, Chongyang Wang, Ruonan He, Weitao Guo, Bo Han, Baoxing Qin, Jiaxin Li
  International Conference on Intelligent Robots and Systems (IROS), October 2022
  [project page] [code] (* denotes equal contributions)

• MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesis
  Jiaxin Li^*, Zijian Feng^*, Qi She, Henghui Ding, Changhu Wang, Gim Hee Lee
  International Conference on Computer Vision (ICCV), October 2021
  [project page] [code] (* denotes equal contributions)

• MT-ORL: Multi-Task Occlusion Relationship Learning
  Panhe Feng, Qi She, Lei Zhu, Jiaxin Li, Lin ZHANG, Zijian Feng, Changhu Wang, Chunpeng Li, Xuejing Kang, Anlong Ming
  International Conference on Computer Vision (ICCV), October 2021
  [code]

• Interaction via bi-directional graph of semantic region affinity for scene parsing
  Henghui Ding, Hui Zhang, Jun Liu, Jiaxin Li, Zijian Feng, Xudong Jiang
  International Conference on Computer Vision (ICCV), October 2021

Challenges

• 1st place (Winner) in DeepFashion2 Cloth Retrieval Challenge
  ICCV DeepFashion2 Workshop, 2019
  Team Hydra@ViSenze

Academic Services

• Conference reviewer: ECCV’22, ICRA’22, CVPR’23, ICCV’23
• Journal reviwer: IJCV