Greetings. My name is Chunhui Liu and I am an applied scientist from Amazon AWS Rekognition Team. I am working on understanding visual concepts in video domain. That contains action/activity understanding, movie/story understanding. I obtained my Master degree in Computer Vision from Carnegie Mellon University. Before that, I obtained my Bachelor degree from Peking University in China.


Experiences

Applied Scientist, Amazon AWS

2020.2 - present

Rekognition Group, Seattle

  • Worked on Spatial Temporal Action Recognition and Detection.
  • Working on Movie Scene Detection.

Research Assistance, Carnegie Mellon University

2019.2 - 2019.7

Human-Computer Interaction Institute

  • Worked on NoRILLA project, a patented mixed-reality educational system bridging physical and virtual worlds to improve STEM learning.

Graduate Student,
Carnegie Mellon University

2018.9 - present

Master of Science in Computer Vision

  • Cumulative QPA: 4.11/4

Research Assistance, Peking University

2016.2 - 2018.7

Wangxuan Institute of Computer Technology

STRUCT Group

  • Worked on Skeleton-Based Action Understanding.

Undergraduate Student, Peking University

2014.9 - 2018.7

Bachelor of Science in Computer Science, Summa Cum Laude

  • GPA:3.59/4.0
  • Outstanding Graduation Award of Peking University
    and Beijing City
  • Outstanding Student Award of Peking University

Publication

(1) Action, Actor, and an uniform way to understand. 2019-present

NUTA: Non-uniform Temporal Aggregation for Action Recognition.
Xinyu Li*, Chunhui Liu*, Bing Shuai, Yi Zhu, Hao Chen, Joseph Tighe

PDF  

A Comprehensive Study of Deep Video Action Recognition .
Yi Zhu, Xinyu Li, Chunhui Liu, Mohammadreza Zolfaghari, Yuanjun Xiong, Chongruo Wu, Zhi Zhang, Joseph Tighe, R. Manmatha, Mu Li

PDF  

Application of Multi-Object Tracking with Siamese Track-RCNN to the Human in Events Dataset .
Bing Shuai, Andrew Berneshawi, Manchen Wang, Chunhui Liu, Davide Modolo, Xinyu Li, Joseph Tighe

Proceedings of the 28th ACM International Conference on Multimedia, 4625-4629  

PDF  

(2) An interesting side-project on interpreting auto encoders. 2017

Patch Correspondences for Interpreting Pixel-level CNNs
Victor Fragoso, Chunhui Liu, Aayush Bansal, Deva Ramanan

Arxiv preprint 1711.10683. PDF

(3) Understanding Human Actions In Skeleton Domain 2016-2018

PKU-MMD: A Large Scale Benchmark for Skeleton-Based Human Action Understanding.
Chunhui Liu, Yueyu Hu, Yanghao Li, Sijie Song, and Jiaying Liu

ACM Multimedia (ACM MM) Workshop, Silicon Valley, California, U.S., Oct. 2017. 

Project  PDF   Code   Dataset

Temporal Perceptive Network for Skeleton-Based Action Recognition.
Yueyu Hu, Chunhui Liu, Yanghao Li, Sijie Song and Jiaying Liu

British Machine Vision Conference (BMVC), London, UK, Sep. 2017. 

PDF

Online Action Detection and Forecast via Multi-Task Deep Recurrent Neural Network.
Chunhui Liu, Yanghao Li, Yueyu Hu and Jiaying Liu

IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP2017), New Orleans, LA, U.S., March. 2017.

Project  PDF


Awards