Guardado en:
Detalles Bibliográficos
Autores principales: Qiu, Bo-Cheng, Lin, Fang-Ying, Sun, Ming-Han, Lin, Yu-Fan, Lee, Chia-Ming, Hsu, Chih-Chung
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2605.22096
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916035411574784
author Qiu, Bo-Cheng
Lin, Fang-Ying
Sun, Ming-Han
Lin, Yu-Fan
Lee, Chia-Ming
Hsu, Chih-Chung
author_facet Qiu, Bo-Cheng
Lin, Fang-Ying
Sun, Ming-Han
Lin, Yu-Fan
Lee, Chia-Ming
Hsu, Chih-Chung
contents Capsule endoscopy event detection is challenging because clinically relevant findings are sparse, visually heterogeneous, and evaluated at the event level rather than by frame accuracy. We propose VISTA, a metric-aligned multi-backbone framework for the RAREVISION task. VISTA combines EndoFM-LV for temporal context and DINOv3 ViTL/16 for frame-level visual semantics, followed by a Diverse Head Ensemble (DHE), Validation-Guided Weighted Fusion (VGWF), and Anatomy-Aware Temporal Event Decoding (ATED). The original official submission achieved hidden-test temporal mAP@0.5 of 0.3530 and mAP@0.95 of 0.3235. After the competition, extending local threshold refinement with a global coarse search improved performance to 0.3726 mAP@0.5 and 0.3431 mAP@0.95, ranking Team ACVLab second in the post-competition evaluation.
format Preprint
id arxiv_https___arxiv_org_abs_2605_22096
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle VISTA: Validation-Guided Integration of Spatial and Temporal Foundation Models with Anatomical Decoding for Rare-Pathology VCE Event Detection -- after competition results
Qiu, Bo-Cheng
Lin, Fang-Ying
Sun, Ming-Han
Lin, Yu-Fan
Lee, Chia-Ming
Hsu, Chih-Chung
Computer Vision and Pattern Recognition
Capsule endoscopy event detection is challenging because clinically relevant findings are sparse, visually heterogeneous, and evaluated at the event level rather than by frame accuracy. We propose VISTA, a metric-aligned multi-backbone framework for the RAREVISION task. VISTA combines EndoFM-LV for temporal context and DINOv3 ViTL/16 for frame-level visual semantics, followed by a Diverse Head Ensemble (DHE), Validation-Guided Weighted Fusion (VGWF), and Anatomy-Aware Temporal Event Decoding (ATED). The original official submission achieved hidden-test temporal mAP@0.5 of 0.3530 and mAP@0.95 of 0.3235. After the competition, extending local threshold refinement with a global coarse search improved performance to 0.3726 mAP@0.5 and 0.3431 mAP@0.95, ranking Team ACVLab second in the post-competition evaluation.
title VISTA: Validation-Guided Integration of Spatial and Temporal Foundation Models with Anatomical Decoding for Rare-Pathology VCE Event Detection -- after competition results
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.22096