Skip to content
AI News HubLIVE
Original source2 min read

VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference

Summary

VigilFormer is a unified framework combining deformable spatio-temporal attention with causal temporal modeling for anomaly detection in untrimmed surveillance video. It achieves AUC scores of 87.83%, 97.21%, and 89.74% on UCF-Crime, ShanghaiTech, and CUHK Avenue respectively, at 41.5 FPS on a single GPU, outperforming recent weakly-supervised methods in both accuracy and speed.

SourcearXiv Computer VisionAuthor: Xinze Zhang
VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[2606.14724] VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference

[Submitted on 31 May 2026]

Title:VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference

View a PDF of the paper titled VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference, by Xinze Zhang

View PDF HTML (experimental)

Abstract:Video anomaly detection in surveillance settings must balance detection accuracy against real-time throughput, a tension that existing methods address either through stronger feature extractors or more efficient architectures, but rarely both. We present VigilFormer, a unified framework that combines deformable spatio-temporal attention with causal temporal modeling to detect anomalies in untrimmed surveillance video. The proposed Deformable Spatio-Temporal Encoder (DSTE) attends to a sparse set of informative locations across frames, avoiding the quadratic cost of dense attention while retaining the ability to capture irregular motion patterns. A Causal Anomaly Classifier (CAC) applies dilated causal convolutions over snippet-level features and optimizes a contrastive multiple-instance learning objective that separates anomalous and normal representations without frame-level labels. To meet deployment constraints, an Adaptive Confidence Scheduler (ACS) dynamically skips low-information frames at inference time, reducing redundant computation in static scenes. Evaluated on UCF-Crime, ShanghaiTech, and CUHK Avenue, VigilFormer achieves AUC scores of 87.83%, 97.21%, and 89.74% respectively, at 41.5 FPS on a single GPU, outperforming recent weakly-supervised methods in both accuracy and speed.

Subjects:

Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)

Cite as: arXiv:2606.14724 [cs.CV]

(or arXiv:2606.14724v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2606.14724

arXiv-issued DOI via DataCite

Submission history

From: Xinze Zhang [view email] [v1] Sun, 31 May 2026 14:27:57 UTC (5,735 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference, by Xinze Zhang

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.CV

new | recent | 2026-06

Change to browse by:

cs cs.AI

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Proposes Deformable Spatio-Temporal Encoder (DSTE) that attends to sparse informative locations to avoid quadratic cost of dense attention while capturing irregular motion.
  • Introduces Causal Anomaly Classifier (CAC) using dilated causal convolutions and contrastive multiple-instance learning to separate anomalous and normal representations without frame-level labels.
  • Adds Adaptive Confidence Scheduler (ACS) to dynamically skip low-information frames during inference, reducing redundant computation in static scenes.
  • Achieves leading AUC scores and real-time throughput of 41.5 FPS on three benchmark datasets.

Highlights and analysis are generated automatically and may contain errors. Check the original source.