Hyun Seok Seong

I'm a postdoctoral researcher at KAIST, working with Prof. Gunhee Kim. I received my Ph.D. in Artificial Intelligence from Sungkyunkwan University, supervised by Prof. Jae-Pil Heo.

My research focuses on image/video representation learning, with applications to visual segmentation/grounding, object-centric learning, and multimodal retrieval. More broadly, I am interested in developing learning frameworks that remain effective under limited supervision. Recently, I've been extending this research toward learning representations of visual objects and their dynamics, with the goal of understanding dynamic environments.

Email / Google Scholar / Github / LinkedIn

Publications

	Dissect and Prune: Enhancing Robustness in AI-Generated Image Detection Dahye Kim, Jaehyun Choi, Hyun Seok Seong, Seongho Kim, Donghun Lee, Sungwon Yi, and JangHo Choi International Conference on Machine Learning (ICML), 2026 Arxiv / Code
	Reconstruction-Guided Slot Curriculum: Addressing Object Over-Fragmentation in Video Object-Centric Learning WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo Computer Vision and Pattern Recognition (CVPR), 2026 Arxiv / Code
	Looking Beyond the Window: Global-Local Aligned CLIP for Training-free Open-Vocabulary Semantic Segmentation ByeongCheol Lee, Hyun Seok Seong, Sangeek Hyun, Gilhan Park, WonJun Moon, and Jae-Pil Heo Computer Vision and Pattern Recognition (CVPR), 2026 Arxiv / Code
	From Vicious to Virtuous Cycles: Synergistic Representation Learning for Video Object-Centric Learning Hyun Seok Seong, WonJun Moon, and Jae-Pil Heo (: equal contribution) International Conference on Learning Representations (ICLR)*, 2026 Arxiv / Code
	Selective Contrastive Learning for Weakly Supervised Affordance Grounding WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo (: equal contribution) International Conference on Computer Vision (ICCV)*, 2025 Arxiv / Code
	Temporal Alignment-Free Video Matching for Few-shot Action Recognition SuBeen Lee, WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo Computer Vision and Pattern Recognition (CVPR), 2025 (Oral presentation) Arxiv / Code
	Foreground-Covering Prototype Generation and Matching for SAM-Aided Few-Shot Segmentation Suho Park, SuBeen Lee, Hyun Seok Seong, Jaejoon Yoo, and Jae-Pil Heo (: equal contribution) AAAI Conference on Artificial Intelligence (AAAI)*, 2025 Arxiv / Code
	Progressive Proxy Anchor Propagation for Unsupervised Semantic Segmentation Hyun Seok Seong, WonJun Moon, SuBeen Lee, and Jae-Pil Heo European Conference on Computer Vision (ECCV), 2024 Arxiv / Code
	Task-Oriented Channel Attention for Fine-Grained Few-Shot Classification SuBeen Lee, WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2024 Arxiv / Code
	Task-disruptive Background Suppression for Few-Shot Segmentation Suho Park, SuBeen Lee, Sangeek Hyun, Hyun Seok Seong, and Jae-Pil Heo AAAI Conference on Artificial Intelligence (AAAI), 2024 Arxiv / Paper / Code
	Leveraging Hidden Positives for Unsupervised Semantic Segmentation Hyun Seok Seong, WonJun Moon, SuBeen Lee, and Jae-Pil Heo Computer Vision and Pattern Recognition (CVPR), 2023 Arxiv / Paper / Code
	Minority-Oriented Vicinity Expansion with Attentive Aggregation for Video Long-Tailed Recognition WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo AAAI Conference on Artificial Intelligence (AAAI), 2023 (Oral presentation) Arxiv / Paper / Code
	TCX: Texture and channel swappings for domain generalization Jaehyun Choi, Hyun Seok Seong, Sanguk Park, and Jae-Pil Heo Pattern Recognition Letters, 2023 Paper
	Difficulty-Aware Simulator for Open Set Recognition WonJun Moon, Junho Park, Hyun Seok Seong, Cheol-Ho Cho, and Jae-Pil Heo European Conference on Computer Vision (ECCV), 2022 Arxiv / Paper / Code
	Pivot-Guided Embedding for Domain Generalization Hyun Seok Seong, Jaehyun Choi, Woojin Jeong, and Jae-Pil Heo IEEE Access, 2022 Paper / Code