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 Video

Updated: Apr 11, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

5.7K

Silhouette analysis for human action recognition based on supervised temporal t-SNE and incremental learning.

Jian Cheng1, Haijun Liu, Feng Wang

  • 1School of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China. justus.cheng@gmail.com

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|June 10, 2015
PubMed
Summary

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

Comparative multidimensional psychosocial profiles of major depressive disorder patients with higher versus lower anhedonia burden: a cross-sectional study.

BMC psychiatry·2026
Same author

Predicting intraoperative blood loss using cervical vertebral bone quality scores in cervical corpectomy and fusion procedure.

European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society·2026
Same author

Brake-Drive Osteo System: Sequential Modulation of the Inflammatory Microenvironment and Osteogenesis for Osteoporotic Bone Defect Regeneration.

Advanced materials (Deerfield Beach, Fla.)·2026
Same author

Preoperative geriatric nutritional risk index is a reliable tool for predicting postoperative adjacent segment disease among elderly patients with degenerative lumbar diseases: a case control study.

BMC musculoskeletal disorders·2026
Same author

Assessment of Bone Mineral Density in Patients With Degenerative Spinal Disease by MRI-based Vertebral Bone Quality Score at Different Lumbar Vertebral Levels: An Observational Prospective Study.

Spine·2026
Same author

Organic-inorganic heterostructure empowers infected wound healing.

Journal of materials chemistry. B·2025

This study introduces supervised temporal t-stochastic neighbor embedding (ST-tSNE) for human action recognition. The method effectively analyzes silhouette sequences using incremental learning for robust performance.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Human action recognition is crucial for surveillance, robotics, and human-computer interaction.
  • Analyzing human silhouette sequences presents challenges due to temporal dynamics and variations.

Purpose of the Study:

  • To develop an effective human action recognition method for silhouette sequences.
  • To enhance action recognition through supervised temporal t-stochastic neighbor embedding (ST-tSNE) and incremental learning.

Main Methods:

  • Supervised temporal t-stochastic neighbor embedding (ST-tSNE) was developed to model relationships between action frames in a manifold.
  • Three incremental learning methods were introduced for low-dimensional embedding of new data: locally linear embedding, locality preserving projection, and manifold-oriented stochastic neighbor projection.

Related Experiment Videos

Last Updated: Apr 11, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

5.7K
  • The methods leverage class label and temporal information for accurate frame representation.

    • Main Results:

    • The proposed ST-tSNE and incremental learning methods demonstrated effectiveness and robustness in human action silhouette analysis.
    • Experimental results showed superior performance compared to existing state-of-the-art methods.
    • The investigated incremental learning techniques successfully preserved the intrinsic action structure in low-dimensional representations.

    Conclusions:

    • The developed ST-tSNE combined with incremental learning offers a powerful approach for human action recognition from silhouette sequences.
    • The proposed methods provide a robust framework for analyzing dynamic human actions in video data.
    • This research contributes to advancing the field of computer vision and pattern recognition.