I am interested in understanding how machine learning systems work and improving them. AI is often inspired by human intelligence, yet they cannot be identical. It is exciting to discover these differences and find ways to leverage them.
Research interests: Large Language Models, Data Bias, Generative AI
Byungju Kim
🔊
Contacts
✉️ byungjukim10@gmail.com✉️ feidfoe@gmail.com
Github/feidfoe
Google Scholar
About
Education
Ph.D in Electric Engineering (Biased Learning, Deep Learning)
KAIST, Korea — 8/2020MS in Electric Engineering (VQA, Deep Learning)
KAIST, Korea — 8/2016BS in Electric Engineering
KAIST, Korea — 8/2014Experience
Research Scientist
Mathpresso Inc — 10/2020 to PresentPublications
Conference
SAAS: Solving Ability Amplification Strategy for Enhanced Mathematical Reasoning in Large Language Models
Hyeonwoo Kim, Gyoungjin Gim, Yungi Kim, Jihoo Kim, Byungju Kim, Wonseok Lee, Chanjun Park
EMNLP (Industry Track), 2024.
[paper]
Patch-Wise Attention Network for Monocular Depth Estimation
Sihaeng Lee, Janghyeon Lee, Byungju Kim, Eojindl Yi, and Junmo Kim
AAAI, 2021.
De-biasing Neural Networks with Estimated Offset for Class Imbalanced Learning
Byungju Kim, Hyeong Gwon Hong, and Junmo Kim
WACV, 2021.
[paper]
Weight Decay Scheduling and Knowledge Distillation for Active Learning
Juseung Yun, Byungju Kim, and Junmo Kim
ECCV, 2020.
[paper]
Collaborative Method for Incremental Learning on Classification and Generation
Byungju Kim, Jaeyoung Lee, Kyungsu Kim, Sungjin Kim and Junmo Kim
ICIP, 2019, p.390-394.
[paper]
[poster]
Learning Not to Learn: Training Deep Neural Networks with Biased Data
Byungju Kim, Hyunwoo Kim, Kyungsu Kim, Sungjin Kim and Junmo Kim
CVPR, 2019, p.9012-9020.
[paper]
[code]
Highway Driving Dataset for Semantic Video Segmentation
Byungju Kim, Junho Yim and Junmo Kim
BMVC, 2018, p.140.
[paper]
Regularization of Deep Network via Latent Subclass Learning
Joochang Kim, Yegang Lee, Byungju Kim, Juseung Yun and Junmo Kim
RiTA, 2017.
Develpment of deep learning-based facial expression recognition system
Heechul Jung, Sihaeng Lee, Sunjeong Park, Byungju Kim, Junmo Kim, Injae Lee, and Chunghyun Ahn
FCV, 2015, doi: 10.1109/FCV.2015.7103729.
[paper]
Journal
Adjusting Decision Boundary for Class Imbalanced Learning
Byungju Kim and Junmo Kim
IEEE Access, vol. 8, pp. 81674-81685, 2020, doi:10.1109/ACCESS.2020.2991231.
[paper]
[code]
Question Aware Prediction with Candidate Answer Recommendation for Visual Question Answering
Byungju Kim and Junmo Kim
Electronics Letters, 2017, 53.18: 1244-1246.
[paper]
Workshop
Data Driven Approach for Mathematical Problem Solving
Byungju Kim, Wonseok Lee, Jaehong Kim and Jungbin Im
MathNLP: The 2nd Workshop on Mathematical Natural Language Processing (co-located with LREC-COLING 2024)
[paper]
Patents
International
Method and Device for Learning Neural Network
Junmo Kim, Byungju Kim, Yeakang Lee, Siheang Lee, Minsuk Park, Pyeonghwan Ahn, Jaeyoung Lee
2018-09-06 / PCT/KR2018/010421
Method and Device for Learning Neural Network for Recognizing Class
Junmo Kim, Byungju Kim, Juchang Kim, Yeakang Lee, Minsuk Park, Juseung Yun, Jaeyoung Lee, Donggyu Ju
2018-09-04 / PCT/KR2018/010271
Domestic (KOR)
Electronic device and operating method for generating caption information for a image sequence
Junmo Kim, Minsuk Park, Byungju Kim, Yeakang Lee, Jaeyoung Lee, Siheang Lee, Kyungsu Kim
2019-05-22 / 10-2019-0060221
Method and Device for Learning Neural Network
Junmo Kim, Byungju Kim, Yeakang Lee, Siheang Lee, Minsuk Park, Pyeonghwan Ahn, Jaeyoung Lee
2017-09-08 / 10-2017-0115464
Method and Device for Learning Neural Network for Recognizing Class
Junmo Kim, Byungju Kim, Juchang Kim, Yeakang Lee, Minsuk Park, Juseung Yun, Jaeyoung Lee, Donggyu Ju
2017-09-08 / 10-2017-0115451
Decorrelation Method using Label Smoothing for Deep Branched Network
Junmo Kim, Byungju Kim, Yeakang Lee, Youngsoo Kim
2017-06-13 / 10-2017-0073860