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

Human action recognition by semilatent topic models.

Yang Wang1, Greg Mori

  • 1Simon Fraser University, Burnaby, Canada. ywang12@cs.sfu.ca

IEEE Transactions on Pattern Analysis and Machine Intelligence
|August 22, 2009
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

Structured Label Inference for Visual Understanding.

IEEE transactions on pattern analysis and machine intelligence·2019
Same author

Deep Neural Network Compression by In-Parallel Pruning-Quantization.

IEEE transactions on pattern analysis and machine intelligence·2018
Same author

Validation of accuracy of SVM-based fall detection system using real-world fall and non-fall datasets.

PloS one·2017
Same author

Multi-Instance Classification by Max-Margin Training of Cardinality-Based Markov Networks.

IEEE transactions on pattern analysis and machine intelligence·2017
Same author

A comparison of accuracy of fall detection algorithms (threshold-based vs. machine learning) using waist-mounted tri-axial accelerometer signals from a comprehensive set of falls and non-fall trials.

Medical & biological engineering & computing·2016
Same author

Distinguishing the causes of falls in humans using an array of wearable tri-axial accelerometers.

Gait & posture·2013
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

We developed new topic models for human action recognition in videos. These models simplify training and improve accuracy by directly linking topics to class labels, outperforming existing methods.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Artificial Intelligence

Background:

  • Human action recognition from video is crucial for applications like surveillance and human-computer interaction.
  • Existing latent topic models for visual recognition face challenges in training complexity and determining the optimal number of topics.

Purpose of the Study:

  • To introduce two novel topic models for enhanced human action recognition from video sequences.
  • To address limitations of previous latent topic models by directly correlating latent topics with class labels and observing latent variables.

Main Methods:

  • A novel "bag-of-words" representation for video sequences, where each frame is treated as a "word."
  • Development of two new topic models with direct correspondence between latent topics and class labels.

Related Experiment Videos

  • Decoupling of model parameters to simplify the training process.

    • Main Results:

    • Achieved significantly improved performance in action classification across five diverse datasets.
    • Demonstrated results comparable to or surpassing previously published benchmarks.
    • Validated the effectiveness of utilizing class label information during training.

    Conclusions:

    • The proposed topic models offer a more efficient and effective approach to human action recognition.
    • Directly incorporating class labels into topic models enhances recognition accuracy and simplifies model selection.
    • These models represent a significant advancement in the field of video-based action recognition.