Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Video behavior profiling for anomaly detection.

Tao Xiang1, Shaogang Gong

  • 1Department of Computer Science, Queen Mary, University of London, London, UK. txiang@dcs.qmul.ac.uk

IEEE Transactions on Pattern Analysis and Machine Intelligence
|March 29, 2008
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Molecular Glues Recruiting RNF213 As an E3 Ligase for Targeted Protein Degradation: A Minimal Dibromoacetamide Warhead As a Recruitment Ligand.

Journal of the American Chemical Society·2026
Same author

Solving the Hubbard model with neural quantum states.

Nature communications·2026
Same author

Pyroptosis-immunity-microbiome axis in acute upper gastrointestinal bleeding: mechanisms, risk prediction, and individualized strategies.

Frontiers in medicine·2026
Same author

The Analgesic Efficacy and Safety of Intramuscular Hydromorphone Versus Butorphanol for Acute Pain in the Emergency Department: A Randomized Trial.

Pain research & management·2026
Same author

Erratum: Possible Observation of Quadrupole Waves in Spin Nematics [Phys. Rev. Lett. 135, 156704 (2025)].

Physical review letters·2026
Same author

Persistent Fermi pockets and robust electron pairing in lightly doped CuO<sub>2</sub> planes of cuprate superconductors.

Nature communications·2026
Same journal

Relation DETR+: Exploring Explicit Position Relation Prior for Dense Prediction.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

RBF++: Quantifying and Optimizing Reasoning Boundaries across Measurable and Unmeasurable Capabilities for Chain-of-Thought Reasoning.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

CAFE: Cross-View Adaptive Fusion and Cluster Center Enhancement for Robust Multi-View Clustering.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Ethics-Aware Safe Reinforcement Learning for Rare-Event Risk Control in Interactive Urban Driving.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Learning Shape Anchors for Holistic Indoor Scene Understanding.

IEEE transactions on pattern analysis and machine intelligence·2026
See all related articles

This study introduces a novel framework for unsupervised surveillance video analysis, enabling accurate normal behavior recognition and anomaly detection without manual labeling. The approach outperforms traditional methods, even with limited or noisy data.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Artificial Intelligence

Background:

  • Surveillance systems require effective methods for recognizing normal behavior and detecting anomalies.
  • Current methods often rely on manually labeled training data, which is time-consuming and costly.

Purpose of the Study:

  • To develop an unsupervised framework for behavior profiling and anomaly detection in surveillance videos.
  • To enable online normal behavior recognition and anomaly detection without manual data labeling.

Main Methods:

  • A novel behavior representation using discrete scene events and Dynamic Bayesian Networks (DBNs).
  • Unsupervised spectral clustering for natural behavior pattern grouping with automatic model and feature selection.
  • A composite generative model for generalizing unseen normal behaviors.

Related Experiment Videos

  • An online Likelihood Ratio Test (LRT) for robust anomaly detection and behavior recognition.

    • Main Results:

    • The proposed framework effectively models video behavior for online recognition and anomaly detection.
    • Unsupervised training yielded superior anomaly detection performance compared to supervised methods on unseen videos.
    • The online LRT method demonstrated advantages over Maximum Likelihood (ML) in differentiating ambiguous behaviors.

    Conclusions:

    • The developed framework offers a robust and efficient solution for unsupervised behavior analysis in surveillance.
    • The approach is effective even with noisy and sparse datasets from diverse indoor and outdoor environments.
    • This unsupervised method provides a significant advancement for real-world anomaly detection applications.