STPrompt \(\boldsymbol{++}\) : Prompting Vision-Language Models for Weakly Supervised Video Anomaly Detection and Fine-Grained Localization

作者
Peng Wu,Chengyu Pan,Guansong Pang,Xiangteng He,Zhiwei Yang,Peng Wang,Yanning Zhang
出处
期刊:ACM Transactions on Multimedia Computing, Communications, and Applications [Association for Computing Machinery]
卷期号:22 (8): 1-22 被引量:1
标识
DOI:10.1145/3828722
摘要

Traditional weakly supervised video anomaly detection (WSVAD) tasks typically rely on coarse-grained frame-level labels for training. Although this approach reduces annotation costs, it results in weak semantic understanding and spatial localization capabilities due to the absence of fine-grained annotations, hindering precise pixel-level anomaly detection and localization. Thanks to the success of vision-language models (VLMs), e.g., CLIP, recent approaches leveraging large VLMs focus on exploiting their strong semantic understanding capabilities, but they typically feed only keyframes or short video segments into the models, without supplying sufficient prior contextual information (e.g., contextual frames around anomalies, zoomed-in anomaly regions, and detailed anomaly descriptions), which restricts the models’ capability for fine-grained anomaly understanding and precise localization. More recently, a few methods leveraging VLMs, attempt to achieve training-free spatial anomaly localization by fusing patch-level visual features with simple textual features. However, these methods employ simplistic textual descriptions, lacking deep semantic comprehension of anomalies, leading to coarse localization results with significant irrelevant background noise. To address these issues, we propose STPrompt \(++\) , a novel weakly supervised spatio-temporal video anomaly detection and localization method based on VLMs. In our work, we systematically leverage preliminary coarse localization regions derived from anomaly scores as spatial priors, together with contextual frames around keyframes, zoomed-in views of suspected anomalous regions, and refined textual descriptions of anomalies. This comprehensive prompting mechanism guides the VLMs toward deep semantic comprehension of video anomalies, enabling accurate pixel-level spatial localization. The proposed STPrompt \(++\) requires no additional training and significantly enhances the precision of anomaly understanding and localization through a carefully designed multi-round and multi-modal prompting mechanism. Extensive experiments on two widely used WSVAD benchmarks, UCF-Crime and UBnormal, show that our method achieves state-of-the-art spatial localization performance and competitive temporal anomaly detection results. Notably, on UCF-Crime dataset, our approach improves spatial localization accuracy (in TIoU) by 5.61% over the current best method (from 23.90% to 29.51%), underscoring its superior capabilities in precise anomaly localization and semantic understanding.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
大个的应助被SunnyZjw采纳,获得10
刚刚
CipherSage的应助被sy采纳,获得10
刚刚
Observer完成签到,获得积分10
1秒前
Code_king发布了新的文献求助10
1秒前
shen完成签到,获得积分10
1秒前
小马甲的应助被可爱觅松采纳,获得10
1秒前
fei完成签到,获得积分20
2秒前
2秒前
ZM发布了新的文献求助10
2秒前
2秒前
周勇军完成签到,获得积分20
2秒前
3秒前
3秒前
3秒前
3秒前
Code_king发布了新的文献求助10
3秒前
Code_king发布了新的文献求助10
3秒前
Code_king发布了新的文献求助10
4秒前
fei发布了新的文献求助10
4秒前
4秒前
领导范儿的应助被黃偉倫采纳,获得10
4秒前
霁月发布了新的文献求助20
4秒前
4秒前
星辰大海的应助被拼搏的青雪采纳,获得10
4秒前
秋风的应助被风趣的笑天采纳,获得10
4秒前
4秒前
Code_king发布了新的文献求助10
5秒前
5秒前
忐忑的琦发布了新的文献求助10
5秒前
5秒前
6秒前
科目三的应助被周勇军采纳,获得10
6秒前
hxh完成签到 ,获得积分20
7秒前
sugar完成签到,获得积分10
7秒前
赵杨关注了科研通微信公众号
7秒前
香蕉觅云的应助被恩雁采纳,获得10
7秒前
荼荼荼发布了新的文献求助10
7秒前
大石发布了新的文献求助10
7秒前
7秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
八旗制度の研究 600
plantTFDB 508
Resource Recovery Handbook 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8175943
求助须知:如何正确求助?哪些是违规求助? 9567501
关于积分的说明 21793186
捐赠科研通 7770490
什么是DOI,文献DOI怎么找? 3431935
关于科研通互助平台的介绍 2556294
邀请新用户注册赠送积分活动 2477485