BK-Lee
Byung-Kwan Lee
School of Electrical Engineering
Ph.D. Candidate, KAIST
Hello! I am Ph.D. candidate in School of Electrical Engineering at Korea Advanced Institute of Science and Technology (KAIST) under the supervision of Prof. Yong Man Ro. I received B.S. for machine learning statistics in Mathematics (major) and for signal & image processing in Electronic Engineering (minor) from Hanyang University at 2018. Also, I received M.S. for developing autonomous vehicle technologies in the CCS graduate school of Mobility from KAIST at 2020. I experienced the team leader for an international autonomous vehicle competition and received the first prize in the president's award at 2018. Now, I have concentrated on building efficient vision language models having super performances for diverse capabilities.
Research Interests (Reliable and Explainable AI based on Self-Supervised Model, Visual Foundation Model, Large Language Model)
Previous: Causal Inference in Robustness and Explainability (e.g., Robust/Non-Robust Feature and Fundamental Reason of Model Vulnerability, and Unsupervised Semantical Clustering by Self-Supervised Learning)
Paper: NeurIPS 2021 [link], CVPR 2022 [link], CVPR 2023 [link], ICCV 2023 [link], ArXiv 2023 [link]
Recent: Building efficient large language and vision models having super performances for diverse capabilities.
Paper: ACL Findings 2024 [link], Preparation for ECCV 2024 [link], Preparation for NeurIPS 2024 [link]
Education
B.S.: Department of Mathematics (Electronic Engineering), Hanyang University, Korea (2014.03~2018.02)
M.S.: The CCS graduate school of Mobility, KAIST, Korea (2018.02~2020.02)
"Thesis: Training encoder-attention through fully-connected CRFs for efficient end-to-end lane detection model [link]"
Ph.D. Candidate: School of Electrical Engineering, KAIST, Korea (2020.03~)
Professional Programming Skills
Python [Pytorch/Tensorflow]
C/C++
Robot Operating System (ROS)
Flutter (Dart) on Android Studio
Publications
Meteor: Mamba-based Traversal of Rationale for Large Language and Vision Models
Byung-Kwan Lee, Chae Won Kim, Beomchan Park, Yong Man Ro
Preparation for NeurIPS 2024 [link][code]
MoAI: Mixture of All Intelligence for Large Language and Vision Models
Byung-Kwan Lee, Beomchan Park, Chae Won Kim, Yong Man Ro
Preparation for ECCV 2024 [link][code]
CoLLaVO: Crayon Large Language and Vision mOdel
Byung-Kwan Lee, Beomchan Park, Chae Won Kim, Yong Man Ro
Findings of the Association for Computational Linguistics (ACL Findings), 2024 [link][code]
Causal Unsupervised Semantic Segmentation
Junho Kim*, Byung-Kwan Lee*, and Yong Man Ro (*: equally contributed)
Mitigating Adversarial Vulnerability through Causal Parameter Estimation by Adversarial Double Machine Learning
Byung-Kwan Lee*, Junho Kim*, and Yong Man Ro (*: equally contributed)
IEEE/CVF International Conference on Computer Vision (ICCV), 2023 [link][code]
Mitigating Dataset Bias in Image Captioning through CLIP Confounder-free Captioning Network
YeonJu Kim, Junho Kim, Byung-Kwan Lee, Sebin Shin, and Yong Man Ro
IEEE International Conference on Image Processing (ICIP), 2023 [link][code]
Demystifying Causal Features on Adversarial Examples and Causal Inoculation for Robust Network by Adversarial Instrumental Variable Regression
Junho Kim*, Byung-Kwan Lee*, and Yong Man Ro (*: equally contributed)
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023 [link][code]
Masking Adversarial Damage: Finding Adversarial Saliency for Robust and Sparse Network
Byung-Kwan Lee*, Junho Kim*, and Yong Man Ro (*: equally contributed)
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022 [link][code]
Distilling Robust and Non-Robust Features in Adversarial Examples by Information Bottleneck
Junho Kim*, Byung-Kwan Lee*, and Yong Man Ro (*: equally contributed)
Neural Information Processing Systems (NeurIPS), 2021 [link][code]
Towards Adversarial Robustness of Bayesian Neural Network through Hierarchical Variational Inference
Byung-Kwan Lee, Youngjoon Yu, and Yong Man Ro
Reviewer Experiences
Journal
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
IEEE Transactions on Neural Networks and Learning Systems (TNNLS)
Elsevier on Computer Vision and Image Understanding (CVIU)
Conference
Computer Vision and Pattern Recognition (CVPR) 2021/2022/2023/2024
International Conference on Computer Vision (ICCV) 2021/2023
European Conference on Computer Vision (ECCV) 2022/2024
International Conference on Learning Representations (ICLR) 2023/2024
Neural Information Processing Systems (NeurIPS) 2023/2024
International Conference on Machine Learning (ICML) 2024
Project Experiences
Research on Human-aligned and User-customized Plug&Play Explainability with Causation for Deep Neural Networks (2022~)
Research on Data-driven Robust Deep Neural Networks against Adversarial Attacks (2020~)
Development of Unmanned Runway Snow Removal Equipment for Unmanned Mobile Robots (2019)
Development of Construction Environment Information by SLAM (Camera-IMU-GPS) for Unmanned Mobile Robots (2019)
Development of Fillet Weld Gap State and Operation Information Recognition Technology Using Deep Neural Networks (2019)
Development of Merging Lane and Object Detection in Real-Time Processing for Self-Driving Technologies (2018~2019)
Development of Autonomous Vehicle Framework in Real-Time Environments by Unifying Operating System (2018~2019)