Beomyoung Kim is an AI research engineer at NAVER's Clova team and a Ph.D. candidate in AI at KAIST, with nine years of experience focused on real-world image understanding. He specializes in image segmentation and object detection, advancing label-efficient methods (weakly/semi-supervised), continual segmentation, and practical deployment such as real-time segmentation on edge devices and human matting. Beomyoung works with vision foundation and multi-modal models like SAM and has hands-on experience adapting architectures for OCR (CRNN with LSTM/GRU) showcased in his GitHub contributions. His profile blends strong academic research under Prof. Sung Ju Hwang with production-minded engineering, often bridging novel algorithms to constrained, real-time systems.
CRNN (CNN+RNN) for OCR using Keras / License Plate Recognition
Role in this project:
ML Engineer
Contributions:37 commits, 2 PRs, 33 pushes in 1 year 7 months
Contributions summary:Beom primarily focused on modifying and updating the model architecture for the CRNN (Convolutional Recurrent Neural Network) for OCR. The commits demonstrate modifications to the `Model.py` and `Model_GRU.py` files, indicating experimentation with different recurrent layers (LSTM and GRU) and architectural changes. Furthermore, the user made changes in `Prediction.py` and `training.py`, suggesting involvement in both the model training and prediction processes.
Contributions:19 commits, 18 pushes, 1 branch in 9 months
license-plateplateinpaintingsyntheticlicense
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.