An Augmented Treble Stream Deep Neural Network for Video Analysis

Chaolong Zhang, Yuanping Xu, Zhijie Xu, Mei Gong, Benjun Guo, Dengguo Yao

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

Abstract

Video analysis for human action recognition is one of the most important research areas in pattern recognition and computer vision due to its wide applications. Deep learning-based approaches have been proven more effective than conventional feature engineering-based models. However, the performance is still unreliable when facing real-world application scenarios. Inspired by the Convolutional Neural Network (CNN) and Recurrent Long-Short Term Model (LSTM), this paper presents an augmented treble-stream deep neural network architecture that supports direct extraction of spatial-Temporal features from video streams and their corresponding dense optical flows. This innovative approach assists effective detection of complex video event features that are annotated by rich event "appearance"and motion features. Substantially improved recognition accuracy is recorded during the experiments that are carried and benchmarked over public video event datasets, for example, UCF 101 and HMDB 51. Analytical evaluation approves the validity and effectiveness of the treble-stream neural network design.

Original languageEnglish
Title of host publication2020 24th International Conference Information Visualisation, IV 2020
EditorsEbad Banissi, Farzad Khosrow-Shahi, Anna Ursyn, Mark W. McK. Bannatyne, Joao Moura Pires, Nuno Datia, Kawa Nazemi, Boris Kovalerchuk, John Counsell, Andrew Agapiou, Zora Vrcelj, Hing-Wah Chau, Mengbi Li, Gehan Nagy, Richard Laing, Rita Francese, Muhammad Sarfraz, Fatma Bouali, Gilles Venturin, Marjan Trutschl, Urska Cvek, Heimo Muller, Minoru Nakayama, Marco Temperini, Tania Di Mascio, Filippo Sciarrone Veronica Rossano Rossano, Ralf Dorner, Loredana Caruccio, Autilia Vitiello, Weidong Huang, Michele Risi, Ugo Erra, Razvan Andonie, Muhammad Aurangzeb Ahmad, Ana Figueiras, Mabule Samuel Mabakane
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages301-306
Number of pages6
ISBN (Electronic)9781728191348
DOIs
Publication statusPublished - Sept 2020
Externally publishedYes
Event24th International Conference Information Visualisation, IV 2020 - Melbourne, Australia
Duration: 7 Sept 202011 Sept 2020

Publication series

NameProceedings of the International Conference on Information Visualisation
Volume2020-September
ISSN (Print)1093-9547

Conference

Conference24th International Conference Information Visualisation, IV 2020
Country/TerritoryAustralia
CityMelbourne
Period7/09/2011/09/20

Keywords

  • CNN and LSTM
  • deep learning
  • human action recognition
  • treble-stream network
  • video analysis

Fingerprint

Dive into the research topics of 'An Augmented Treble Stream Deep Neural Network for Video Analysis'. Together they form a unique fingerprint.

Cite this