Bo Fu
Applied Scientist II at Amazon Fulfillment Technologies & Robotics
Greater Boston United States
Summary
🤩
Rockstar🎓
Top SchoolBo Fu is a Machine Learning Engineer with nine years of experience focused on improving reproducibility and usability of transfer learning methods. He contributes to prominent open-source efforts like thuml's Transfer-Learning-Library, standardizing training scripts and seed initialization across domain adaptation algorithms such as MCD, DAN, and CDAN. Skilled at debugging and harmonizing code formats, he brings practical engineering rigor to research-grade ML codebases. Based in China, he blends hands-on implementation with attention to experiment reliability—a detail often overlooked in ML engineering. His GitHub activity highlights a pragmatic approach: making advanced domain adaptation methods easier to reproduce and adopt.
10 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy (Ph.D.) Robotics, Doctor of Philosophy (Ph.D.) Robotics at University of MichiganBachelor of Engineering Vehicle Engineering (Automotive Electronics), Bachelor of Engineering Vehicle Engineering (Automotive Electronics) at Tongji UniversityMaster of Science Mechanical Engineering, Master of Science Mechanical Engineering at Carnegie Mellon University
Chinese, English, German
Github Skills (4)
deeplearning-ai10
python10
deep-learning10
pytorch8
Programming languages (4)
Github contributions (5)
thuml/Transfer-Learning-Library
Apr 2020 - May 2021
Transfer Learning Library for Domain Adaptation, Task Adaptation, and Domain Generalization
Role in this project:
ML Engineer
ML EngineerContributions:24 commits, 3 PRs, 14 pushes in 1 year 1 month
Contributions summary:Bo's commits primarily involve updating and modifying training scripts and example code within the transfer learning library. They focused on standardizing and incorporating seed initialization across various domain adaptation methods, including MCD, DAN, and CDAN. The user also addressed format inconsistencies and corrected errors within the code. These changes suggest a focus on improving the reproducibility and usability of different transfer learning algorithms.
transfer-learningdomain-adaptationfinetuneunsupervised-domain-adaptationimage-translation
thuml/Calibrated-Multiple-Uncertainties
Jul 2020 - Sep 2021
Code Release for "Learning to Detect Open Classes for Universal Domain Adaptation"(ECCV2020)
Contributions:5 commits, 4 pushes, 1 branch in 1 year 2 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.