Guanbin Huang
Computer Vision Research Engineer at DeepBlue Technology
Songjiang District, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolGuanbin Huang is a computer vision research engineer with eight years of experience, currently applying advanced AI techniques at DeepBlue Technology in Shanghai. He holds a Distinction MSc in Artificial Intelligence from Queen Mary University of London and a BE in Mechatronics and Robotics, combining strong academic foundations with hands-on engineering. Guanbin has practical expertise optimizing detection models for production, notably enhancing CenterNet and adapting DCNv2 and custom CUDA kernels to run efficiently with TensorRT. His background spans both technical delivery and cross-functional coordination from a prior project management role at L'Oréal, giving him a knack for translating research into operational monitoring and deployment. Colleagues can reach him via GitHub and WeChat (peterdeep) for collaboration on model acceleration and inference engineering.
8 years of coding experience
Master's degree, Artificial Intelligence, 78(Distinction), Master's degree, Artificial Intelligence, 78(Distinction) at Queen Mary University of LondonBachelor of Engineering - BE, Mechatronics, Robotics, and Automation Engineering, 3.3/4.0, Bachelor of Engineering - BE, Mechatronics, Robotics, and Automation Engineering, 3.3/4.0 at Shanghai University of Engineering Science
Github Skills (11)
tensorrt10
cuda10
pytorch10
c-language10
deep-learning10
cprogramming-language10
python10
object-detection9
computer-vision9
faster-rcnn8
mask-rcnn8
Programming languages (5)
Github contributions (5)
shouxieai/tensorRT_Pro
Sep 2021 - Nov 2021
C++ library based on tensorrt integration
Role in this project:
ML Engineer
ML EngineerContributions:38 commits, 32 pushes, 5 comments in 1 month
Contributions summary:Guanbin contributed significantly to the development and enhancement of the CenterNet model within the TensorRT integration project. Their work includes updating and modifying Python and C++ code for pre- and post-processing steps. Furthermore, the user adjusted the DCNv2 implementation and improved the CUDA kernel for object detection, specifically NMS (Non-Maximum Suppression). The updates reflect efforts to make the model compatible with TensorRT.
cppc-libraryyolov7tensorflowtensorrt
Guanbin-Huang/camera_calibration_cpp
Oct 2021 - Aug 2022
Contributions:4 commits, 2 pushes in 10 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.