FS-SegDiff: Mitigating Intra-Class Variation and Attention Leakage for Robust Diffusion-Based Few-Shot Semantic Segmentation

Few-shot semantic segmentation (FSS) has gained a significant attention as a promising solution to segmenting the target objects of unseen classes using only a few annotated examples. Recently, pretrained diffusion models have emerged as an effective approach for FSS, acting as both a powerful featu...

ver descrição completa

Detalhes bibliográficos

Principais autores:	Jaehyun Kim, Wonyong Seo, Munchurl Kim
Formato:	Artigo
Idioma:	Inglês
Publicado em:	IEEE 2026-01-01
Colecção:	IEEE Access
Assuntos:	Attention leakage diffusion diffusion-based few-shot semantic segmentation few-shot semantic segmentation multi-scale feature matching
Acesso em linha:	https://ieeexplore.ieee.org/document/11357902/
Tags:	Adicionar Tag Sem tags, seja o primeiro a adicionar uma tag!