Yuhan Shen - Applied Scientist at Amazon

Yuhan Shen

Applied Scientist at Amazon

Boston, Massachusetts, United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Yuhan Shen is an Applied Scientist at Amazon AGI with a Ph.D. in Computer Science from Northeastern University, specializing in video understanding and multimodal learning. Over eight years of experience spans top-tier research internships at Meta FAIR and TikTok, producing CVPR publications and leading oral and workshop papers on weakly-supervised video segmentation and large-scale audio-visual captioning. Their PhD work focused on interpreting complex human activities from video with minimal supervision, yielding multiple first-author CVPR papers and measurable gains in segmentation IoU. Yuhan also contributes to open-source ML tooling, improving backend performance and reservoir-sampling logic in PaddlePaddle projects, signaling strong systems and engineering chops beyond pure research. Based in Boston, they bridge rigorous academic research with production-aware engineering to move multimodal models from prototype to deployable components. A less obvious strength is the blend of deep theoretical grounding and hands-on backend optimization, enabling both novel algorithms and scalable implementation.

8 years of coding experience

Doctor of Philosophy - PhD, Computer Science, 4.0, Doctor of Philosophy - PhD, Computer Science, 4.0 at Northeastern University

Bachelor's degree, Electrical and Electronics Engineering, Bachelor's degree, Electrical and Electronics Engineering at Tsinghua University

Github Skills (14)

deep-learning10

python10

data-structure9

performance-optimization9

package-management9

file-management9

software-packaging9

management9

data-structures9

configuration-management9

paddlepaddle9

visualization8

visualizations8

modello7

Programming languages (5)

TypeScriptC++VueHTMLPython

Github contributions (5)

PaddlePaddle/PaddleFormers

Sep 2019 - Jan 2021

PaddleFormers is an easy-to-use library of pre-trained large language model zoo based on PaddlePaddle.

Role in this project:

Back-end Developer

Contributions:8 reviews, 178 commits, 172 PRs in 1 year 4 months

Contributions summary:Yuhan primarily contributed to the configuration and management of the PaddleHub environment. Their commits focused on adding and modifying config operators, including reset, log level setting, and server URL configuration. Furthermore, the user implemented features related to module installation and file management, such as generating an MD5 checksum for installed modules.

large-language-modelspaddlepaddlemodel

PaddlePaddle/VisualDL

Mar 2020 - Jun 2021

Deep Learning Visualization Toolkit（『飞桨』深度学习可视化工具）

Role in this project:

Back-end Developer

Contributions:21 reviews, 95 commits, 156 PRs in 1 year 3 months

Contributions summary:Yuhan primarily contributed to the `visualdl/server/data_manager.py` file, implementing features related to data management, specifically focusing on accumulating sampling rates. They fixed performance issues related to backend data transfer. The code changes involved modifications to data structures and algorithms within the reservoir sampling context, indicating backend logic development for the visualization toolkit.

deep-learningvisualizationpaddlepaddleonnxcaffe

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.

Request Free Trial