About Me

I am a second-year PhD student in Mathis Lab at EPFL, under the supervision of professor Mackenzie Mathis. My research interest is to bridge neurosience and AI. In the AI side, I am building more generalizable animal pose estimator that allows across species, across datasets inference. While such a task acts as a good playground to study deep nets’ generalization and adaptivitiy, it also helps to provide better tools for neuroscientists to study animals’ behavior. At the same time, I also work on using realistic musculoskeletal model and neural networks to reverse engineer the brain of laboratory mice.


  • [Mar 2022] our paper about multi animal pose estimation and tracking was accepted by Nature Methods.

  • [Mar 2022] our paper about building more robust vision transformer was accpted by CVPR 2022.

  • [Mar 2021] 2 papers about adversarial vulnerability of neural networks were accpeted by CVPR 2021.

  • [Dec 2021] I passed my candidacy exam.

  • [Dec 2020] I moved to Geneva and joined Mathis Lab as a PhD student.

Work experience:

  • I worked as a senior algorithm engineer at Alibaba Group.

  • I worked as a researcher at IIISCT.