VISTA: Validation-Guided Integration of Spatial and Temporal Foundation Models with Anatomical Decoding for Rare-Pathology VCE Event Detection -- after competition results

TL;DR

VISTA is a novel framework for rare-pathology capsule endoscopy event detection that combines multiple models and validation techniques, achieving high accuracy and ranking second in a competition.

Contribution

The paper introduces VISTA, a multi-backbone, validation-guided framework that effectively integrates spatial and temporal models for improved rare-pathology event detection.

Findings

Achieved a post-competition mAP@0.5 of 0.3726

Achieved a post-competition mAP@0.95 of 0.3431

Ranked second in the competition evaluation

Abstract

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.