Feng Cheng
- I am currently a Master student at Shanghai Jiao Tong Univerisity, and I'm expected to graduate in March, 2019. My research advisor is Dr. Shi-Lin Wang.
- I received my Bachelor's degree in Shanghai Jiao Tong University.
- My primary research interests include spatial-temporal sequence learning, lip content recognition and authentication.
I am skilled in c/c++, python, java. Beyond these, I can quickly transfer to a new language or new area if needed (they are almost the same except for the rules). I have written several projects in several languages actually:
- A c library jpeg_toolbox_python with both c API and python API. The library is to extract the DCT coefficient of the JPEG images.
- My implementation of paper LipNet: End-to-End Sentence-level Lipreading using Keras(Tensorflow backend). Interestingly, this project got 16 stars and 7 forks.
- An Android app which implements the idea proposed in my paperVisual speaker authentication with random prompt texts by a dual-task CNN framework. We perform authentication with liveness detection only with visual information, specifically, the lip region. The code has not been open-sourced due to business limits. click here for video demo
Publications
[1] Feng Cheng, Shi-Lin Wang, and Alan Wee-Chung Liew. Visual speaker authentication with random prompt texts by a dual-task CNN framework. Pattern Recognition 83 (2018): 340-352.
[2] Xizi Wang, Feng Cheng, Shi-Lin Wang, et al. Adult Image Classification by a Local-Context Aware Network. In 2018 25th IEEE International Conference on Image Processing (ICIP) (pp. 2989-2993). IEEE, 2018.
[3] Feng Cheng, Shi-Lin Wang, Xizi Wang, et al. A global and local context integratin DCNN for adult image classification. In submission to Pattern Recognition.
Honors
- 2016 First prize of the Ninth National Information Security Competition
- 2016 Second Prize of the 13th Huawei Cup National Graduate Mathematical Modeling Contest
- 2017 Wenyuan Pan scholarship (7000 RMB)
- 2018 National Cyberspace safety scholarship(50000 RMB)
Copyright © 2015 Powered by MWeb, Theme used GitHub CSS.