Min-hung Chen
Staff Research Scientist at NVIDIA
New Taipei, Taiwan
Summary
🤩
Rockstar🎓
Top SchoolMin-hung Chen is a Staff Research Scientist at NVIDIA Research with a decade of experience advancing multimodal and efficient AI, particularly in vision-language, 4D/spatial understanding, and transformer-based systems. He holds a Ph.D. from Georgia Tech and has an extensive publication record including ICML (oral), NeurIPS, ICLR, CVPR and ICCV contributions, reflecting both theoretical and applied strengths. Prior roles at Microsoft and MediaTek and internships at Baidu and PlayStation show a track record of translating research into deployable solutions for facial liveness, edge Transformers, and video understanding. He actively shapes the field as an area chair and workshop organizer for top conferences, and his open-source work includes core contributions to temporal CNN/LSTM pipelines for activity recognition. Colleagues describe him as someone who blends rigorous domain-adaptation and self-supervised methods with practical engineering to push multimodal systems toward real-world robustness.
10 years of coding experience
11 years of employment as a software developer
Master of Science (MS), Electrical and Electronics Engineering, Master of Science (MS), Electrical and Electronics Engineering at National Taiwan UniversityDoctor of Philosophy (Ph.D.), Electrical and Computer Engineering, Doctor of Philosophy (Ph.D.), Electrical and Computer Engineering at Georgia Institute of Technology
Chinese, English, Japanese
Github Skills (7)
neural-network10
machine-learning10
convolutional-neural-networks10
torch-net9
torch-hub9
video-understanding8
python4
Programming languages (6)
Github contributions (5)
chihyaoma/Activity-Recognition-with-CNN-and-RNN
May 2016 - Jun 2020
Temporal Segments LSTM and Temporal-Inception for Activity Recognition
Role in this project:
ML Engineer
ML EngineerContributions:182 commits, 152 pushes, 1 branch in 4 years 2 months
Contributions summary:Min-hung implemented a CNN spatial encoder, suggesting contributions to the core machine learning model within the activity recognition project. Further development was done in the TCNN (Temporal Convolutional Neural Network) codes, showing an involvement in a crucial component of the system. The user's work is centered around integrating and enhancing CNN and TCNN aspects to recognize activities based on temporal and spatial information.
pytorchlstm-neural-networksdeep-learninglstmrecognition
cmhungsteve/SSTDA
Mar 2020 - Feb 2021
[CVPR 2020] Action Segmentation with Joint Self-Supervised Temporal Domain Adaptation (PyTorch)
Contributions:19 commits, 2 PRs, 14 pushes in 11 months
pytorchsuperviseddeep-learningcvpr-2020temporal
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.