Sungmin Cha (차성민), Ph.D.
Sungmin Cha (차성민), Ph.D.
Embedded Files
Assistant Professor/Faculty Fellow
Assistant Professor/Faculty Fellow
CS Dept. at the Courant Institute of Mathematical Sciences
CS Dept. at the Courant Institute of Mathematical Sciences
New York University
New York University
CV / Github / Google Scholar
/ Email : sungmin dot cha at nyu dot edu
Hi, I am Sungmin Cha who is working as a Faculty Fellow at NYU, under the supervision of Prof. Kyunghyun Cho. My research have focused on developing adaptive AI models that can learn and evolve efficiently. I’m particularly passionate about Continual Learning—designing cost-efficient algorithms that enable neural networks to learn sequentially without forgetting, much like how humans acquire knowledge over time. Along the way, I’ve also explored Machine Unlearning and efficient post-training strategy for Large Language Models (LLMs), always seeking new challenges and insights at the intersection of adaptive and trustworthy AI.
Previously, I earned my Ph.D. in Electrical and Computer Engineering from Seoul National University (SNU) under the supervision of Prof. Taesup Moon. I also obtained my Master's and Bachelor's degrees in Information and Communication Engineering and Computer Engineering from DGIST and Pukyong National University, respectively. During my Ph.D. studies, I had the opportunity to work as a visiting researcher at Harvard University, as well as a research scientist intern at both NAVER AI and LG AI Research. Additionally, I have been honored with several awards, including the Qualcomm Innovation Fellowship Korea (2021), the Yulchon AI START Fellowship (2022), the Distinguished Doctoral Dissertation Award from the Department of ECE at SNU (2023), and the Best Dissertation Award from the S-OIL Foundation and the Korean Academy of Science and Technology (2024), among others.
Ongoing Research Topics
Ongoing Research Topics
Continual learning (CL)
- continual pre-training, continual instruction tuning for LLMs.
- cost-efficient CL algorithms.
- appropriate evaluation protocol for CL.
- CL algorithms and scenarios in diverse domains (e.g., anomaly detection, long-tailed classification using CLIP, etc.).Machine unlearning (MU)
- MU algorithms for LLMs and classification models.
- appropriate evaluation methods for MU for LLMs and classification models.
News
News
[2025. 06. 10] One paper was accepted to ICML 2025 Workshop on Machine Unlearning for Generative AI!
[2025. 05. 19] One paper was accepted to Interspeech 2025!
[2025. 04. 25] One paper was accepted to TMLR!
[2025. 03. 05] One paper was accepted to ICLR 2025 Workshop Weight Space Learning!
[2025. 01. 22] One paper was accepted to ICLR 2025!
[2024. 12. 23] I gave an invited talk at Hanyang Univ. and Korea Univ.!
[2024. 12. 07] I gave an invited talk at SeoulTech, Seoul National Univ., and Sogang Univ.!
[2024. 12. 05] I received the Best Dissertation Award from the S-OIL and the Korean Academy of Science and Technology.
[2024. 10. 10] One paper was accepted to Workshop on Adaptive Foundation Models, NeurIPS 2024!
[2024. 09. 23] I gave an invited talk at CAU-Core AI Tech Seminar.
Publications (* co-first author, ^ corresponding author)
Publications (* co-first author, ^ corresponding author)
Preprints
Preprints
[P9] Reference-Specific Unlearning Metrics Can Hide the Truth: A Reality Check
Sungjun Cho, Dasol Hwang, Frederic Sala, Sangheum Hwang, Kyunghyun Cho, and Sungmin Cha^
Workshop on Machine Unlearning for Generative AI, ICML 2025
[P8] Why Knowledge Distillation Works in Generative Models: A Minimal Working Explanation
Sungmin Cha and Kyunghyun Cho
[P7] Continual-Mega: A Large-scale Benchmark for Generalizable Continual Anomaly Detection
Geonu Lee, Yujeong Oh, Geonhui Jang, Soyoung Lee, Jeonghyo Song, Sungmin Cha, and YoungJoon Yoo
[P6] Toward Practical Class-Incremental Learning: Long-Tailed Adaptation using CLIP
Hyeokjun Lee, Sungmin Cha and Youngjoon Yoo
[P5] Toward a Holistic Approach To Continual Model Merging
Hoang Phan, Sungmin Cha, Tung Lam Tran, and Qi Lei
[P4] Bridging Few-Shot and Zero-Shot Anomaly Detection: Toward Generalizable Anomaly Detection
Geonu Lee, Soyoung Lee, Jeonghyo Song, Sungmin Cha, and YoungJoon Yoo
[P3] Memory Is Not the Bottleneck: Cost-Efficient Continual Learning via Weight Space Consolidation
Dong Kyu Cho, Taesup Moon, Rumi Chunara, Kyunghyun Cho, and Sungmin Cha^
Workshop on Weight Space Learning, ICLR 2025
[P2] Are We Truly Forgetting? A Critical Re-examination of Machine Unlearning Evaluation Protocols
Yongwoo Kim*, Sungmin Cha* and Donghyun Kim
[P1] Improving Generative Pre-Training: An In-depth Study of Masked Image Modeling and Denoising Models
Hyesong Choi, Daeun Kim, Sungmin Cha, Kwang Moo Yi, and Dongbo Min
Conference Papers
Conference Papers
[C20] Cross-Modal Watermarking for Authentic Audio Recovery and Tamper Localization in Synthesized Audiovisual Forgeries
Minyoung Kim, Youngseo Kim, Sungmin Cha, and Paul Hongsuck Seo
Interspeech, August 2025
[C19] Towards Robust and Parameter-Efficient Knowledge Unlearning for Large Language Models [Code]
Sungmin Cha*, Sungjun Cho*, Dasol Hwang, and Moontae Lee
Workshop on Adaptive Foundation Models, NeurIPS 2024
International Conference on Learning Representation (ICLR), April 2025
[C18] Salience-Based Adaptive Masking: Revisiting Token Dynamics for Enhanced Pre-training [Code]
Hyesong Choi, Hyejin Park, Kwang Moo Yi, Sungmin Cha, and Dongbo Min
The 18th European Conference on Computer Vision (ECCV), October 2024
[C17] Regularizing with Pseudo-Negatives for Continual Self-Supervised Learning [Code]
Sungmin Cha, Kyunghyun Cho, and Taesup Moon
The Forty-first International Conference on Machine Learning (ICML), July 2024
Workshop track at CoLLAs 2024
[C16] Towards Realistic Incremental Scenario in Class Incremental Semantic Segmentation [Code]
Jihwan Kwak, Sungmin Cha, and Taesup Moon
Third Conference on Lifelong Learning Agents (CoLLAs), July 2024
[C15] Towards More Diverse Evaluation of Class Incremental Learning: A Representation Learning Perspective
Sungmin Cha, Jihwan Kwak, Dongsup Shim, Hyunwoo Kim, Moontae Lee, Honglak Lee, and Taesup Moon
Third Conference on Lifelong Learning Agents (CoLLAs), July 2024
Workshop on Continual Learning (CLVision), CVPR 2023
[C14] Learning to Unlearn: Instance-wise Unlearning for Pre-trained Classifiers [Code]
Sungmin Cha*, Sungjun Cho*, Dasol Hwang*, Honglak Lee, Taesup Moon, and Moontae Lee
The Thirty-Eight AAAI Conference on Artificial Intelligence (AAAI), February 2024
[C13] NCIS: Neural Contextual Iterative Smoothing for Purifying Adversarial Perturbations
Sungmin Cha, Naeun Ko, Heewoong Choi, Youngjoon Yoo, and Taesup Moon
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Jan 2024
The best paper award (first award) at CKAIA 2021
[C12] Knowledge Unlearning for Mitigating Privacy Risks in Language Models [Code]
Joel Jang, Dongkeun Yoon, Sohee Yang, Sungmin Cha, Moontae Lee, Lajanugen Logeswaran, and Minjoon Seo
The 61st Annual Meeting of the Association for Computational Linguistics (ACL), July 2023
[C11] Rebalancing Batch Normalization for Exemplar-based Class Incremental Learning [Code]
Sungmin Cha, Sungjun Cho, Dasol Hwang, Soonwon Hong, Moontae Lee, and Taesup Moon
Conference on Computer Vision and Pattern Recognition (CVPR), June 2023
Workshop on Continual Learning (CLVision, 3rd Edition) at CVPR 2022
[C10] SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning [Code]
Sungmin Cha*, Beomyoung Kim*, Youngjoon Yoo, and Taesup Moon (* equal contribution)
Proceedings of Neural Information Processing Systems (NeurIPS), December 2021
International Workshop on Continual Semi-Supervised Learning at IJCAI 2021
The best poster presentation award (first award) at AIIS 2021 Fall Retreat, Nov 2021
[C9] FBI-Denoiser: Fast Blind Image Denoiser for Poisson-Gaussian Noise (Oral) [Code]
Jaeseok Byun*, Sungmin Cha*, and Taesup Moon (* equal contribution)
Conference on Computer Vision and Pattern Recognition (CVPR), July 2021
The best poster presentation award (third award) at AIIS 2021 Spring Retreat, April 2021
Qualcomm Innovation Fellowship Korea (QIFK) 2021
[C8] CPR: Classifier-Projection Regularization for Continual Learning [Code]
Sungmin Cha, Hsiang Hsu, Taebaek Hwang, Flavio P. Calmon, and Taesup Moon
International Conference on Learning Representation (ICLR), May 2021
4th Lifelong Learning Workshop at ICML 2020
The best paper award (first award) from Microsoft Research at CKAIA 2020
Qualcomm Innovation Fellowship Korea (QIFK) 2021
[C7] GAN2GAN: Generative noise learning for blind image denoising with single noisy images [Code]
Sungmin Cha, Taeeon Park, Byeongjoon Kim, Jongduck Baek, and Taesup Moon
International Conference on Learning Representation (ICLR), May 2021
Deep Inverse Workshop at NeurIPS 2020 (Oral)
[C6] Continual Learning with Node-Importance based Adaptive Group Sparse Regularization [Code]
Sangwon Jung*, Hongjoon Ahn*, Sungmin Cha, and Taesup Moon (* equal contribution)
Proceedings of Neural Information Processing Systems (NeurIPS), December 2020
Workshop on Continual Learning in Computer Vision at CVPR 2020
Qualcomm Innovation Fellowship Korea (QIFK) 2020
[C5] Uncertainty-based continual learning with adaptive regularization [Code]
Hongjoon Ahn*, Sungmin Cha*, Donggyu Lee, and Taesup Moon (* equal contribution)
Proceedings of Neural Information Processing Systems (NeurIPS), December 2019
[C4] Fully Convolutional Pixel Adaptive Image Denoiser [Code(Keras)]
Sungmin cha and Taesup Moon
International Conference on Computer Vision (ICCV), November 2019
[C3] DoPAMINE: Double-sided Masked CNN for Pixel Adaptive Multiplicative Noise Despeckling (Oral) [Code]
Sunghwan Joo, Sungmin Cha, and Taesup Moon
The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI), February 2019
[C2] UDLR Convolutional Network for Adaptive Image Denoiser
Sungmin Cha and Taesup Moon
6th International Conference on Robot Intelligence Technology and Applications (RiTA), December 2018
[C1] Neural adaptive image denoiser
Sungmin Cha and Taesup Moon
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), April 2018
Journal Papers
Journal Papers
[J3] Hyperparameters in Continual Learning: A Reality Check
Sungmin Cha and Kyunghyun Cho
Transactions on Machine Learning Research (TMLR), April 2024
Workshop on Continual Learning (CLVision), CVPR 2024
Workshop track at CoLLAs 2024
[J2] Observations on K-image Expansion of Image-Mixing Augmentation for Classification
Joonhyun Jeong*, Sungmin Cha*, Youngjoon Yoo, Sangdoo Yun, Taesup Moon, and Jongwon Choi (* equal contribution)
IEEE Access (IF=3.476), Feb 2023
Synthetic Data Generation Workshop at ICLR 2021
The best paper award (excellence award) at JKAIA 2021
[J1] Interpreting machine learning models in neuroimaging: Towards a unified framework
Lada Kohoutova, Juyeon Heo, Sungmin Cha, Taesup Moon, Tor D. Wager and Choong-Wan Woo
Nature Protocols (IF=11.334), March 2020
Ph.D. Thesis
Ph.D. Thesis
Novel Loss, Layer, and Applications for Neural Network based Continual Learning
Advisor: Taesup Moon (SNU)
Committees: Bohyung Han, Kyomin Jung, Jonghyun Choi (SNU), and Hyunwoo Kim (Korea Univ)
Distinguished Dissertation Award from ECE, SNU
Teaching
Teaching
[2023 Fall, 2024 Spring, 2024 Fall] Data Structures, CS department at CIMS, New York University
Education
Education
Ph.D. [2018.3 ~ 2023.08] (Advisor : Taesup Moon)
Thesis: Novel Loss, Layer, and Applications for Neural Network based Continual Learning
Department of Electrical and Computer Engineering
Seoul National University (SNU)
(Moved from Sungkyunkwan University (SKKU) after Sep 2021)
MS [2016.3 ~ 2018.2] (Advisor : Taesup Moon and Suha Kwak)
Department of Information and Communication Engineering
Daegu Gyeongbuk Institute of Science and Technology (DGIST)
BS [2009.3 ~ 2016.2]
Department of Computer Engineering
Professional Experience
Professional Experience
2023.09 ~ Current: Assistant Professor/Faculty Fellow in CS department at the CIMS, NYU (advisor: Kyunghyun Cho).
2020.12 ~ 2021.11: Research intern, LG AI Research (advisor: Moontae Lee)
2020.11 ~ 2021.05: Research intern, NAVER AI Lab (advisor: YoungJoon Yoo)
2020.01 ~ 2020.07: Visiting scholar, Harvard SEAS (advisor : Flavio Calmon)
2021.03 ~ 2023.08: Research assistant, M.IN.D Lab (advisor : Taesup Moon), SNU
2018.03 ~ 2021.02: Research assistant, M.IN.D Lab (advisor : Taesup Moon), SKKU
2016.03 ~ 2018.02: Research assistant, M.IN.D Lab (advisor : Taesup Moon), DGIST
2014.01 ~ 2015.12: Undergraduate research assistant, Embedded system Lab, PKNU
2010.01 ~ 2011.11: Discharged upon completing military service as a sergeant, Republic of Korea Marine Crops
Award and Scholarship
Award and Scholarship
Award
Award
2024: The Best Dissertation Award from S-OIL Foundation and Korean Academy of Science and Technology
2024: Travel Award (Financial Aid Scholarship), CoLLAs 2024
2024 : Travel Award (Doctoral Consortium), WACV 2024
2023 : Distinguished Dissertation Award from the Department of Electrical and Computer Engineering, SNU
2022 : Second Prize, best poster presentation award, AIIS 2022 Spring Retreat
2021 : First Prize, best poster presentation award, AIIS 2021 Fall Retreat
2021 : Excellence Prize, best paper award, JKAIA 2021
2021 : First Prize, best paper award from COMIPO, CKAIA 2021
2020 : First Prize, best paper award from Microsoft Research, JKAIA 2020
2020 : First Prize, award of superior research, SKKU ICC
2019 : First Prize, best poster presentation award, Samsung AI Forum 2019
2015 : First Prize, Samsung & PKNU software contest, PKNU
2014 : First Prize, Team project competition, PKNU
Scholarship
Scholarship
2022 : Youlchon AI STAR Fellowship
2021 : Qualcomm Innovation Fellowship Korea (QIFK) 2021 (for two papers!)
2021 : Scholarship from department of electrical and computer engineering, SNU
2021 : Academic excellence scholarship, SNU
2020 : Scholarship for long-term overseas training (for six months), BK(Brain Korea)21+, SKKU
2018 ~ 2020 : BK(Brain Korea)21+ Scholarship, SKKU
2016 ~ 2017 : Full scholarship, DGIST
2014 ~ 2015 : Samsung SCSC mentoring scholarship, PKNU
2012 ~ 2015 : Academic excellence scholarship (full scholarship for six semesters), PKNU
2009 : Admission scholarship (for one semester), PKNU
Academic Services
Academic Services
Conference Area Chair
: Continual FoMo Workshop (2024 - present)Conference Reviewer
: NeurIPS (2020 - present), CVPR (2021 - present), CoLLAs (2023 - present), ACL (2024 - present), ICLR (2024 - present),
CVPR CLVISION (2021 - present), U&Me Workshop (2024 - present), ICML (2021), ICCV (2021), ECCV (2022), CVPR NTIRE (2020)Journal Reviewer
: TPAMI (2022 - present)
Invited Presentation
Invited Presentation
2024.12 : Oral presentation at SeoulTech, Seoul National Univ., Sogang Univ., Hanyang Univ., and Korea Univ.
2024.09 : Oral presentation, CAU-Core AI Tech Seminar
2024.05 : Oral presentation, Techart Conference 2024, CAU
2024.01 : Oral presentation, Doctoral Consortium, WACV 2024
2023.09 : Oral presentation, CILVR Seminar, New York University
2022.08 : Oral presentation, SNU AI Summer School 2022, Seoul National University
2022.08 : Oral presentation, KAIA2022, Jaejoo Ireland
2022.07 : Oral presentation, HY-Vision Lab, Hanyang University
2021.11 : Oral presentation, Qualcomm Innovation Fellowship Korea (QIFK) 2021
2021.11 : Poster presentation, AIIS 2021 Fall Retreat
2021.11 : Oral presentation, Open Seminar at LG AI Research
2021.04 : Poster presentation, AIIS 2021 Spring Retreat
2021.07 : Oral presentation, CKAIA 2021
2020.11 : Oral presentation, JKAIA 2021
2020.11 : Oral presentation, Qualcomm Innovation Fellowship Korea (QIFK) 2020
2019.10 : Poster presentation, Samsung AI Forum 2019
2018.10 : Poster presentation, DGIF 2018
Mentoring Experiences (while pursuing Ph.D.)
Mentoring Experiences (while pursuing Ph.D.)
Sunghwan Joo (PhD student at M.IN.D Lab, SKKU)
Donggyu Lee (PhD student at M.IN.D Lab, SKKU)
Hongjoon Ahn (MS student at M.IN.D Lab, SKKU => PhD student at M.IN.D Lab, SNU)
Taeeon Park (MS student at M.IN.D Lab, SKKU => PhD student at M.IN.D Lab, SNU)
Jaeseok Byun (MS student at M.IN.D Lab, SKKU => PhD student at M.IN.D Lab, SNU)
Taebaek Hwang (BS student at SKKU => Currently at Dealicious)
Saehwan Kim (BS student at SNU)
Hyungseo Ahn (BS student at SNU)
Soonwon Hong (BS student at SNU)
Heewoong Choi (BS student at SNU => PhD student at M.IN.D Lab, SNU)
Youngin Kim (BS student at SNU)
Dongwook Lee (BS student at SNU)
Jihwan Kwak (PhD student at M.IN.D Lab, SNU)
Previous News
Previous News
[2024. 07. 01] One paper was accepted to ECCV 2024!
[2024. 06. 26] Two papers were accepted to a workshop track at CoLLAs 2024!
[2024. 05. 28] I received the financial aid scholarship to attend CoLLAS 2024!
[2024. 05. 24] I gave an invited talk at TechArt Conference 2024, CAU.
[2024. 05. 01] One paper was accepted to ICML 2024!
[2024. 04. 26] Two papers were accepted to CoLLAs 2024!
[2024. 03. 29] One paper was accepted to Workshop on Continual Learning (CLVision), CVPR 2024!
[2023. 12. 09] One paper was accepted to AAAI 2024!
[2023. 11. 11] The application of the WACV 2024 Doctoral Consortium was accepted!
[2023. 10. 24] One paper was accepted to WACV 2024!
[2023. 09. 28] I gave an invited talk at CILVR seminar, New York University.
[2023. 09. 01] I started a new chapter as an Assistant Professor/Faculty Fellow in the Computer Science department at the Courant Institute of Mathematical Sciences, New York University, working with Prof. Kyunghyun Cho!
[2023. 07. 31] I won the Distinguished Ph.D. Dissertation Award from the Department of Electrical and Computer Engineering, SNU.[2023. 06. 29] I successfully completed my Ph.D. defense!
[2023. 05. 02] One paper was accepted to ACL 2023!
[2023. 02. 28] One paper was accepted to CVPR 2023!
[2023. 01. 30] One paper was accepted to IEEE Access!
[2022. 08. 07] I gave a poster presentation for one paper at KCCV 2022, Coex
[2022. 08. 03] I gave an invited talk for one paper at SNU AI Summer School 2022, Seoul National University.
[2022. 08. 02] I gave an invited talk for one paper at KAIA2022, Jaejoo Ireland.
[2022. 07. 13] I received Youlchon AI STAR Fellowship.
[2022. 07. 08] I gave an invited talk for Continual Learning at HY-Vision Lab, Hanyang University.
[2022. 04. 15] I gave a poster presentation for one paper at AIIS 2022 Spring Retreat and received second prize in the best presentation award!
[2022. 04. 12] One paper was accepted to Workshop on Continual Learning in Computer Vision at CVPR 2022.
[2022. 01. 27] One paper got an award (Future Gauss Lecturer) from Gauss Labs!
[2021. 12. 01] I started a research internship at LG AI Research (advisor: Moontae Lee).
[2021. 11. 22] Two papers won Qualcomm Innovation Fellowship Korea (QIFK) 2021 and I am nominated as the best voter! [article]
[2021. 11. 19] I gave a poster presentation for one paper at AIIS 2021 Fall Retreat and received first prize in the best presentation award!
[2021. 11. 04] One paper got first prize in the best paper award of JKAIA 2021.
[2021. 11. 04] I gave an invited talk at the open seminar of LG AI Research.
[2021. 09. 29] One paper was accepted to NeurIPS 2021.
[2021. 07. 21] One paper was accepted to International Workshop on Continual Semi-Supervised Learning at IJCAI 2021.
[2021. 07. 09] One paper got best paper award from KOMIPO at CKAIA 2021.
[2021. 06. 22] One paper was accepted to Adversarial Machine Learning Workshop at ICML 2021.
[2021. 04. 30] My three papers are presented at AIIS 2021 Spring Retreat and one paper won the best presentation award (third award).
[2021. 04. 01] One paper was accepted to Synthetic Data Generation Workshop at ICLR 2021.
[2021. 03. 02] I transferred my PhD course to Seoul National University by following my advisor.
[2021. 03. 01] One paper was accepted to CVPR 2021 as oral presentation.
[2021. 01. 13] Two papers were accepted to ICLR 2021.
[2020. 12. 16] One paper got the best paper award from Microsoft Research at JKAIA 2020. [article]
[2020. 12. 11] One paper got the best paper award at Qualcomm Innovation Fellowship Korea 2020.
[2020. 11. 30] I started a research internship at NAVER AI Lab (advisor: YoungJoon Yoo).
[2020. 10. 24] One paper was accepted to NeurIPS 2020 Deep Inverse Workshop as oral presentation.
[2020. 09. 26] One paper was accepted to NeurIPS 2020.
[2020. 06. 25] One paper was accepted to 4th Lifelong Learning Workshop at ICML 2020.
[2020. 04. 14] One paper was accepted to the CVPR 2020 Workshop on Continual Learning!
[2020. 03. 18] One paper was published in Nature Protocols (IF=11.334).
[2020. 02. 17] I got the winner of the First Prize in the award of the Superior Research from SKKU ICC.
[2020. 02. 03] I started to work as a visiting scholar at Harvard SEAS (Advisor : Prof. Flavio Calmon).
[2019. 11. 05] I gave a poster presentation at Samsung AI Forum 2019 and I got the Frist Prize as the best poster presentation.
[2019. 09. 03] One paper was accepted to NeurIPS 2019.
[2019. 07. 23] One paper was accepted to ICCV 2019.
[2018. 11. 01] One paper was accepted to AAAI 2019 as an oral presentation.
[2018. 02. 15] One paper was accepted to ICASSP 2018.
Page updated
Google Sites
Report abuse