Robert Wang
Principal AI Scientist at Teledyne FLIR
Oakville, Ontario, Canada
Summary
👤
Senior🎓
Top SchoolRobert Wang is a Principal AI Scientist with over a decade of experience building high-performance on-device AI, computer vision, and deep learning systems for embedded and UAV platforms. He authored Pelee/Pelee2—real-time CNN detectors (one patented and cited for mobile object detection) that deliver state-of-the-art speed-accuracy trade-offs and are deployed across FLIR business units. At Teledyne FLIR he advanced cross-platform inference engines, distributed training, and runtime optimizations (pruning, quantization, TensorRT/CUDA) to enable solutions like few-shot re-id, vSLAM, and mono depth that run multiple times faster than peers. Previously he led ML engineering for healthcare AI services, building fast distributed data pipelines and mobile/cloud APIs, and earlier won Huawei’s Gold Award for a global distributed OSS. Based in Oakville, Canada, he combines production-grade systems engineering (C++, CUDA, Python) with research-driven innovation and active contributions to mobile object-detection tooling.
11 years of coding experience
13 years of employment as a software developer
Bachelor of Science Computer Science, Bachelor of Science Computer Science at Wuhan UniversityMaster of Science (M.Sc.) Computer Science, Master of Science (M.Sc.) Computer Science at Western UniversityNanodegree, Nanodegree at Udacity
English, Chinese
Github Skills (9)
object-detection10
computer-vision10
machine-learning10
eval10
caffe10
trainings10
python10
evaluation10
modeling10
Programming languages (2)
Github contributions (5)
Robert-JunWang/Pelee
Jan 2018 - Jan 2019
Pelee: A Real-Time Object Detection System on Mobile Devices
Role in this project:
ML Engineer
ML EngineerContributions:53 commits, 1 PR, 47 pushes in 11 months
Contributions summary:Robert contributed code related to training and evaluating a real-time object detection system. This included implementing an evaluation script, `eval_voc.py`, and integrating layers for classification, particularly within the `peleenet.py` file. Further contributions involved adding code to evaluate a single image using the trained model and merging tools for the model architecture.
mobileobject-detection
Robert-JunWang/PeleeNet
Mar 2018 - Feb 2019
PeleeNet: An efficient DenseNet architecture for mobile devices
Contributions:9 commits, 16 pushes, 2 branches in 11 months
densenetmobilemobile-devicesdeep-learningefficient-algorithm
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.