Skip to search boxSkip to navigationSkip to main content

A novel audio-oriented learning strategies for character recognition

  • Changbin Lu
    ,
  • Guangyu Gao(corresponding author)
*Corresponding author for this work
Output:
Chapter in Book/Report/Conference proceeding
Conference contribution
Academic
Peer-review

Information

Output type

Output:
Chapter in Book/Report/Conference proceeding
Conference contribution
Academic
Peer-review

Original language

English

Article number

7938239

Pages from-to (Number of pages)

Pages 459-464 (6 pages)

Publication milestones

  • Published - 2016

Publication status

Published - 2016

Place of publication

Piscataway

Publisher

Institute of Electrical and Electronics Engineers, United States
978-1-5090-5189-2

ISBN (Electronic)

978-1-5090-5188-5

Host publication title

Proceedings - 2016 International Conference on Virtual Reality and Visualization, ICVRV 2016

Host publication editors

  • Dandan Ding
  • Dangxiao Wang
  • Jian Chen
  • Xun Luo

Funding Details

National Natural Science Foundation of China under Grant NO. 61401023

Related Event

Title

6th International Conference on Virtual Reality and Visualization (ICVRV 2016)

Event type

Conference

Date

24/09/2016 - 26/09/2016

Location

Hangzhou, ZhejiangChina

Abstract

<p>In this paper, we propose a robust audio-oriented learning strategies to address the issue of character recognition in movie/TV-series. Identifying major characters in movies/TV-series has drawn researcher's great interests. Most of them have explored some character recognition and retrieval applications based on visual appearance, whereas visual appearance is inconsistent throughout the whole video. Our approach, mainly focusing on audio, features that: (i) we extract both spectral and temporal audio features of Mel-scale Frequency Cepstral Coefficients(MFCC), prosodic, average pause length, speaking rate features, pitch and short time energy, and also the complementarity of Gabor features, (ii) we adopt Multi-Task Joint Sparse Representation and Recognition (MTJSRC) model for learning with all the features except Gabor, and SVM model with Gabor features, (iii) regarding these original features as seeds, we extend the training set from talk shows with semi-supervise learning, (iv) the Conditional Random Field (CRF) model with consideration of the constrains in time sequence is introduced to enhance the final labelling. Finally, experimental results demonstrates the effectiveness performance of our approach.</p>

Metrics

Metrics

SciVal
Author count
2
SciVal
Paper percentile
28