Roy Tseng
R&D Hardware Engineer at Broadcom
Irvine, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRoy Tseng is an R&D hardware engineer focused on transistor-level analog and digital IC design, currently working at Broadcom after internships that included multiport register file work targeting 2nm processes. Based in Irvine and educated at UC Santa Barbara, he combines hands-on PCB and control hardware experience from Rockwell Automation with research exposure at OPUS Lab. With a decade of industry involvement and a parallel background in computer vision and deep learning—evidenced by contributions to a Detectron.pytorch implementation—he bridges hardware design and ML-aware software tooling. Roy is driven to build intuition across disciplines, routinely diving into circuit-level details while remaining fluent in software frameworks used for modern vision research. Colleagues describe him as a curious engineer who prefers implementing practical fixes and features that make complex systems measurable and robust.
10 years of coding experience
1 year of employment as a software developer
Bachelor of Science - BS, Electrical Engineering, Bachelor of Science - BS, Electrical Engineering at UC Santa Barbara
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (6)
mask-rcnn10
computer-vision10
pytorch10
deep-learning10
back-end-development9
faster-rcnn9
Programming languages (10)
Github contributions (5)
roytseng-tw/Detectron.pytorch
Feb 2018 - Apr 2019
A pytorch implementation of Detectron. Both training from scratch and inferring directly from pretrained Detectron weights are available.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:233 commits, 6 PRs, 109 pushes in 1 year 2 months
Contributions summary:Roy primarily contributed to the development of the Detectron.pytorch framework, specifically focusing on modifications related to the Mask R-CNN architecture. Their work included updating training code, fixing bugs in mask loss computation, and implementing features for augmented testing. They also made changes to the RPN and FPN components, likely to adapt the framework for different backbones and training configurations.
pytorchweightstrainingpytorch-implementationdetectron
roytseng-tw/adversarial-pose-pytorch
Dec 2017 - Jan 2018
Contributions:48 commits, 34 pushes in 1 month
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.