đź‘‹ About Me
I am an integrated M.S.–Ph.D. researcher in Artificial Intelligence at Korea University, South Korea. Previously, I received my B.Eng. in Internet of Things Engineering from Hohai University, China.
My research connects Physical AI, brain–computer interfaces, and explainable AI. I study how intelligent systems can learn from human behavior and neural signals while remaining understandable to the people who supervise them. In particular, I work on:
- Physical AI and neuroadaptive oversight — embodied agents that learn from implicit human feedback and adapt visual, auditory, and tactile guidance for safer human–AI collaboration.
- Brain–computer interfaces — EEG decoding, cross-subject adaptation, and multimodal interfaces for motor imagery, tactile perception, emotion recognition, and neural–language modeling.
- Explainable AI — counterfactual explanations that identify object- and part-level evidence behind fine-grained misclassification.
I welcome international research collaboration on neuroadaptive embodied systems, EEG representation learning, and trustworthy machine learning. Feel free to email me or explore my Google Scholar.
🔥 News
- 2026: Our paper has been accepted at NeurIPS 2026.
- 2026: Our paper was selected as Best Student Paper at SMC 2026.
- 2026: NeuroLex was accepted to the International BCI Conference 2026.
- 2025: Our work on fine-grained counterfactual explanations appeared at CVPR 2025.
- 2024: GITGAN was published in Pattern Recognition.
Selected Publications
* indicates equal contribution.
NeurIPS 2026Semantic-Level Invariant Representation Learning for Cross-Hospital Clinical EEG Modeling, Kang Yin, Hye-Bin Shin, Seong-Whan LeeBCI 2026NeuroLex: A Lightweight Domain Language Model for EEG Report Understanding and Generation, Kang Yin, Hye-Bin ShinBCI 2025EEG-based Multimodal Representation Learning for Emotion Recognition, Kang Yin, Hye-Bin Shin, Dan Li, Seong-Whan LeeBCI 2024Pioneering EEG motor imagery classification through counterfactual analysis, Kang Yin, Hye-Bin Shin, Hee-Dong KimPattern Recognition 2024GITGAN: Generative Inter-subject Transfer for EEG Motor Imagery Analysis, Kang Yin, Elissa Yanting Lim, Seong-Whan LeeBCI 2023Target-centered subject transfer framework for EEG data augmentation, Kang Yin, Byeong-Hoo Lee, Byoung-Hee Kwon, Jeong-Hyun ChoSMC 2026Best Student PaperDecoding Error-Related Potentials under Multisensory Feedback with Varying Congruency, Yixin Liu, Kang Yin, Hye-Bin Shin, Seong-Whan LeeBCI 2026Cross-Modal Correspondence Improves ErrP Decoding in Multimodal BCIs, Yixin Liu, Kang Yin, Hye-Bin ShinCVPR 2025Towards Fine-Grained Interpretability: Counterfactual Explanations for Misclassification with Saliency Partition, Lintong Zhang*, Kang Yin*, Seong-Whan LeeNeural Networks 2025Semantic Prioritization in Visual Counterfactual Explanations with Weighted Segmentation and Auto-Adaptive Region Selection, Lintong Zhang, Kang Yin, Seong-Whan LeeBCI 2025Towards a Network Expansion Approach for Reliable Brain-Computer Interface, Byeong-Hoo Lee, Kang YinEMBC 2024Baseline-Guided Representation Learning for Noise-Robust EEG Signal Classification, Elissa Yanting Lim, Kang Yin, Hye-Bin Shin, Seong-Whan LeeBCI 2024Sparse Multitask Learning for Efficient Neural Representation of Motor Imagery and Execution, Hye-Bin Shin, Kang Yin, Seong-Whan LeeBCI 2026Cross-Modal Consistency-Guided Active Learning for Affective BCI Systems, Hyo-Jeong Jang, Hye-Bin Shin, Kang YinIEEE TNSRE 2025Reconstructing unseen sentences from speech-related biosignals for open-vocabulary neural communication, Deok-Seon Kim, Seo-Hyun Lee, Kang Yin, Seong-Whan LeeBCI 2025Personalized continual EEG decoding: Retaining and transferring knowledge, Dan Li, Hye-Bin Shin, Kang YinEMBC 2024Domain-Incremental Learning Framework for Continual Motor Imagery EEG Classification Task, Dan Li, Hye-Bin Shin, Kang Yin, Seong-Whan LeeBCI 2024Personalized continual EEG decoding framework for knowledge retention and transfer, Dan Li, Hye-Bin Shin, Kang Yin
Featured Projects
Research systems that turn neural decoding, embodied intelligence, and adaptive interaction into working experiments.
Belief-Aware Oracle Particle Dynamics
A belief-aware, object-centric dynamics model that learns multiple plausible 3D futures from structured object-state histories.
Neuroadaptive Oversight for Embodied Agents
A closed-loop human–AI system that decodes implicit neural error signals and adapts multimodal cues for safer embodied autonomy.
MetaSkin Neural–Tactile Interface
An immersive VR/AR platform for decoding multi-site tactile EEG and recognizing continuous tactile input in real time.
Motor Imagery Robotic Arm Control
A VR-based four-class motor-imagery BCI with real-time EEG decoding and robotic arm interaction across users.
đź“– Education
- 2021.03–2027.02: Korea University — Integrated M.S. & Ph.D. in Artificial Intelligence.
- 2016.09–2020.06: Hohai University — B.Eng. in Internet of Things Engineering.
