Decoupled Sensitivity-Consistency Learning for Weakly Supervised Video Anomaly Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zheng, Hantao, Han, Ning, Zeng, Yawen, Chen, Hao
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918399958843392
author Zheng, Hantao
Han, Ning
Zeng, Yawen
Chen, Hao
author_facet Zheng, Hantao
Han, Ning
Zeng, Yawen
Chen, Hao
contents Recent weakly supervised video anomaly detection methods have achieved significant advances by employing unified frameworks for joint optimization. However, this paradigm is limited by a fundamental sensitivity-stability trade-off, as the conflicting objectives for detecting transient and sustained anomalies lead to either fragmented predictions or over-smoothed responses. To address this limitation, we propose DeSC, a novel Decoupled Sensitivity-Consistency framework that trains two specialized streams using distinct optimization strategies. The temporal sensitivity stream adopts an aggressive optimization strategy to capture high-frequency abrupt changes, whereas the semantic consistency stream applies robust constraints to maintain long-term coherence and reduce noise. Their complementary strengths are fused through a collaborative inference mechanism that reduces individual biases and produces balanced predictions. Extensive experiments demonstrate that DeSC establishes new state-of-the-art performance by achieving 89.37% AUC on UCF-Crime (+1.29%) and 87.18% AP on XD-Violence (+2.22%). Code is available at https://github.com/imzht/DeSC.
format Preprint
id arxiv_https___arxiv_org_abs_2603_19780
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Decoupled Sensitivity-Consistency Learning for Weakly Supervised Video Anomaly Detection
Zheng, Hantao
Han, Ning
Zeng, Yawen
Chen, Hao
Computer Vision and Pattern Recognition
I.4.8; I.4.9
Recent weakly supervised video anomaly detection methods have achieved significant advances by employing unified frameworks for joint optimization. However, this paradigm is limited by a fundamental sensitivity-stability trade-off, as the conflicting objectives for detecting transient and sustained anomalies lead to either fragmented predictions or over-smoothed responses. To address this limitation, we propose DeSC, a novel Decoupled Sensitivity-Consistency framework that trains two specialized streams using distinct optimization strategies. The temporal sensitivity stream adopts an aggressive optimization strategy to capture high-frequency abrupt changes, whereas the semantic consistency stream applies robust constraints to maintain long-term coherence and reduce noise. Their complementary strengths are fused through a collaborative inference mechanism that reduces individual biases and produces balanced predictions. Extensive experiments demonstrate that DeSC establishes new state-of-the-art performance by achieving 89.37% AUC on UCF-Crime (+1.29%) and 87.18% AP on XD-Violence (+2.22%). Code is available at https://github.com/imzht/DeSC.
title Decoupled Sensitivity-Consistency Learning for Weakly Supervised Video Anomaly Detection
topic Computer Vision and Pattern Recognition
I.4.8; I.4.9
url https://arxiv.org/abs/2603.19780