Published in SLEEP, 2025
This paper is about automatic sleep staging AI model. We developed a lightweight AI model using knowledge distillation, while providing verifiability.
Recommended citation: Park, K., Hong, J., Lee, W., Shin, H.W. and Kim, H.S., 2025. DistillSleep: Real-Time, On-Device, Interpretable Sleep Staging from Single-Channel EEG. SLEEPJ, p.zsaf240.
Download Paper | Download Bibtex
Published in ICLR, 2026
T1 is a CNN-Transformer hybrid that binds channels to attention heads for robust time series imputation, achieving 46% better performance than existing methods, especially under extreme missingness.
Recommended citation: Park, D., Ryu, H., Bae, S., Park. K., Kim. H.S., 2026., T1: One-to-One Channel-Head Binding for Multivariate Time-Series Imputation., The Fourteenth International Conference on Learning Representations.
Download Paper
Published in NeurIPS, 2025
This position paper introduces a new direction for AI development: Adaptive sensing
Recommended citation: Baek, E., Park, K., Ko, J., Oh, M.H., Gong, T. and Kim, H.S., 2025. AI Should Sense Better, Not Just Scale Bigger: Adaptive Sensing as a Paradigm Shift. The Thirty-Ninth Annual Conference on Neural Information Processing Systems Position Paper Track.
Download Paper
Published in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024
This paper is about covariate shift of AI models, caused by changes in environment or sensor parameters.
Recommended citation: Baek, E., Park, K., Kim, J. and Kim, H.S., 2024. Unexplored faces of robustness and out-of-distribution: Covariate shifts in environment and sensor domains. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 22294-22303).
Download Paper | Download Slides | Watch on YouTube
Published in International Conference on Information Processing in Sensor Networks, 2023
This paper is about optimizing on-device 3D object detection on mobile devices equipped with CPU-GPU-NPU.
Recommended citation: Park, K., Choi, Y.R., Lee, I. and Kim, H.S., 2023, May. PointSplit: Towards on-device 3D object detection with heterogeneous low-power accelerators. In Proceedings of the 22nd International Conference on Information Processing in Sensor Networks (pp. 67-81).
Download Paper | Download Slides | Watch on YouTube
Published in Arxiv, 2022
This paper proposes a federated semi-supervesed learning method, by sharing knowledge between cleints and server through prototypes.
Recommended citation: Kim, W., Park, K., Sohn, K., Shu, R. and Kim, H.S., 2022. Federated semi-supervised learning with prototypical networks. arXiv preprint arXiv:2205.13921.
Download Paper
Published in Arxiv, 2020
This paper introduces a new application for real-time mask detection on-device, utilizing Google Coral EdgeTPU
Recommended citation: Park, K., Jang, W., Lee, W., Nam, K., Seong, K., Chai, K. and Li, W.S., 2020. Real-time mask detection on google edge TPU. arXiv preprint arXiv:2010.04427.
Download Paper