- 게시일자
- 조회수
- 297
박진선 교수 연구실(VIPLab), BK21 우수 국제학술대회 ACM MM 논문 1편 발표
박진선 교수 연구실(VIPLab), BK21 우수 국제학술대회 ACM MM 논문 1편 발표
부산대학교 시각 지능 및 인지 연구실(VIPLab)이 2025년 10월 27일~31일 아일랜드 더블린에서 개최된 BK21 CS분야 우수 국제학술대회(BK IF: 4)인 ACM Multimedia (ACM MM) 2025에 1편의 논문(Regular Paper)을 발표하였다.
시각 지능 및 인지 연구실(VIPLab) 석박통합과정 조용현 (제1저자), 지도교수 박진선, 박사과정 김장현 (제2저자) (좌측부터)
제목: BAC-GCN: Background-Aware CLIP-GCN Framework for Unsupervised Multi-Label Classification
저자: 조용현, 김장현, 박진선
연구요약: Multi-label classification has recently demonstrated promising performance through CLIP-based unsupervised learning. However, existing CLIP-based approaches primarily focus on x-object-centric features, which limits their ability to capture rich contextual dependencies between x-objects and their surrounding scenes. In addition, the vision transformer architecture of CLIP exhibits a bias toward the most prominent x-object, often failing to recognize small or less conspicuous x-objects precisely. To address these limitations, we propose Background-Aware CLIP-GCN (BAC-GCN), a novel framework that explicitly models class-background interactions and is designed to capture fine-grained visual patterns of small x-objects effectively. BAC-GCN is composed of three key components: (i) a Similarity Kernel that extracts patch-level local features for each category (i.e., class and background), (ii) a CLIP-GCN that captures relational dependencies between local-global and class-background features, and (iii) a Re-Training for Small x-objects (ReSO) strategy that enhances the representation of small and hard-to-learn x-objects by learning their distinctive visual characteristics. Therefore, our method facilitates a deeper understanding of complex visual contexts, enabling the model to make decisions by leveraging diverse visual cues and their contextual relationships. Extensive experiments demonstrate that BAC-GCN achieves state-of-the-art performance on three benchmark multi-label datasets: VOC07, COCO, and NUS, validating the effectiveness of our approach. The project page is available at: https://github.com/yonghyeonjo46/BAC-GCN
주요그림:
- 첨부파일
- 첨부파일이(가) 없습니다.