Semantic-Deviation-Anchored Multi-Branch Fusion for Unsupervised Anomaly Detection and Localization in Unstructured Conveyor-Belt Coal Scenes

TL;DR

This paper introduces a new benchmark, CoalAD, for unsupervised anomaly detection in coal conveyor scenes, and proposes a multi-branch fusion framework that improves detection and localization accuracy in complex, unstructured environments.

Contribution

The paper presents CoalAD, a novel benchmark dataset, and a multi-branch fusion method that leverages complementary cues for robust anomaly detection and localization in unstructured coal scenes.

Findings

Our method outperforms baseline approaches on CoalAD.

Ablation studies confirm the effectiveness of each component.

The framework achieves high accuracy in pixel-level localization.

Abstract

Reliable foreign-object anomaly detection and pixel-level localization in conveyor-belt coal scenes are essential for safe and intelligent mining operations. This task is particularly challenging due to the highly unstructured environment: coal and gangue are randomly piled, backgrounds are complex and variable, and foreign objects often exhibit low contrast, deformation, occlusion, resulting in coupling with their surroundings. These characteristics weaken the stability and regularity assumptions that many anomaly detection methods rely on in structured industrial settings, leading to notable performance degradation. To support evaluation and comparison in this setting, we construct \textbf{CoalAD}, a benchmark for unsupervised foreign-object anomaly detection with pixel-level localization in coal-stream scenes. We further propose a complementary-cue collaborative perception framework that extracts and fuses complementary anomaly evidence from three perspectives: object-level semantic composition modeling, semantic-attribution-based global deviation analysis, and fine-grained texture matching. The fused outputs provide robust image-level anomaly scoring and accurate pixel-level localization. Experiments on CoalAD demonstrate that our method outperforms widely used baselines across the evaluated image-level and pixel-level metrics, and ablation studies validate the contribution of each component. The code is available at https://github.com/xjpp2016/USAD.