Spatiotemporal features learning with 3DPyraNet

Research output: Chapter in Book or Conference Publication/ProceedingConference Publicationpeer-review

5 Citations (Scopus)

Abstract

A discriminative approach based on the 3DPyraNet model for spatiotemporal feature learning is proposed. In combination with a linear SVM classifier, our model outperform state-of-the-art methods on two datasets (KTH, Weizmann). Whereas, shows comparable result with current best methods on third dataset (YUPENN). The features are compact, achieving 94.08 %, 99.13 %, and 94.67% accuracy on KTH, Weizmann, and YUPENN, respectively. The proposed model appears more suitable for spatiotemporal feature learning compared to traditional feature learning techniques; also, the number of parameters is far less than other 3DConvNets.

Original languageEnglish
Title of host publicationAdvanced Concepts for Intelligent Vision Systems - 17th International Conference, ACIVS 2016, Proceedings
EditorsCosimo Distante, Dan Popescu, Paul Scheunders, Wilfried Philips, Jacques Blanc-Talon
PublisherSpringer-Verlag
Pages638-647
Number of pages10
ISBN (Print)9783319486796
DOIs
Publication statusPublished - 2016
Externally publishedYes
Event17th International Conference on Advanced Concepts for Intelligent Vision Systems, ACIVS 2016 - Lecce, Italy
Duration: 24 Oct 201627 Oct 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10016 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference17th International Conference on Advanced Concepts for Intelligent Vision Systems, ACIVS 2016
Country/TerritoryItaly
CityLecce
Period24/10/1627/10/16

Keywords

  • Action recognition
  • Deep learning
  • Dynamic scene understanding
  • Pyramidal neural network

Fingerprint

Dive into the research topics of 'Spatiotemporal features learning with 3DPyraNet'. Together they form a unique fingerprint.

Cite this