Yaodong WANG
Senior Software Engineer at WeChat- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
Credentials
-
CUDA Programming
英伟达Dec, 2020- Nov, 2024
Experience
-
WeChat
-
China
-
Technology, Information and Internet
-
1 - 100 Employee
-
Senior Software Engineer
-
Sep 2021 - Present
WeChat WeChat
-
-
-
Tencent
-
China
-
Software Development
-
700 & Above Employee
-
Software Engineer
-
Dec 2019 - Sep 2021
Improve the performance of training procedure among multiple GPUs. Using profiling tools to locate bottlenecks of program and accelerate it.
-
-
Software Engineer Intern
-
Jun 2019 - Nov 2019
Focused on optimization of TensorFlow. Made contributions to TensorFlow on Github.
-
-
-
British Council
-
United Kingdom
-
Non-profit Organizations
-
700 & Above Employee
-
Exam Service Officer
-
Jun 2017 - Dec 2020
-
-
-
Lenovo
-
China
-
IT Services and IT Consulting
-
700 & Above Employee
-
Lenovo H.K. Lab Intern
-
Dec 2018 - Mar 2019
Implement large scale distributed system for Lenovo Machine Intelligence Platform Implement large scale distributed system for Lenovo Machine Intelligence Platform
-
-
-
The University of Hong Kong
-
Hong Kong
-
Higher Education
-
700 & Above Employee
-
Research Program Assistant
-
Sep 2018 - Dec 2018
Following Prof. C.L. Wang, working on distributed training of deep neural networks on large scale clusters and workers. Our research group introduce a new parameter server design, which solves problems in memory useage and synchronization Following Prof. C.L. Wang, working on distributed training of deep neural networks on large scale clusters and workers. Our research group introduce a new parameter server design, which solves problems in memory useage and synchronization
-
-
-
Tencent
-
China
-
Software Development
-
700 & Above Employee
-
AI-Lab Internship
-
Apr 2018 - Aug 2018
Following Dr. Wenhan Luo, I engaged in a Silent Face Anti-spoof Project. Goal: to develop algorithm for silent face anti-spoof (face liveness detection without user cooperation), and deploy it to Tencent products. · 99% positive pass rate, and nearly 100% rejection rate of negative samples. In real-world products, success rate of engine call is improved from 50% to 70%, success rate of user call is improved from 85% to 90%. · Built a Django based web service. · Main tools & frames: Tensorflow, Caffe, C++, Python. Show less
-
-
Education
-
The University of Hong Kong
Master of Science - MS, Computer Science -
Rutgers University–New Brunswick
Exchange Student, Electronic Engineering -
South China University of Technology
Bachelor of Engineering - BE, Electronic & Information Engineering -
South China University of Technology
Master's degree, Computer Science