Welcome to the InfoSci Platform
Reference Hub1
Optical Flow Prediction for Blind and Non-Blind Video Error Concealment Using Deep Neural Networks

Optical Flow Prediction for Blind and Non-Blind Video Error Concealment Using Deep Neural Networks

Arun Sankisa, Arjun Punjabi, Aggelos K. Katsaggelos
Copyright: © 2019 |Volume: 10 |Issue: 3 |Pages: 20
ISSN: 1947-8534|EISSN: 1947-8542|EISBN13: 9781522565345|DOI: 10.4018/IJMDEM.2019070102
Cite Article Cite Article

MLA

Sankisa, Arun, et al. "Optical Flow Prediction for Blind and Non-Blind Video Error Concealment Using Deep Neural Networks." IJMDEM vol.10, no.3 2019: pp.27-46. http://doi.org/10.4018/IJMDEM.2019070102

APA

Sankisa, A., Punjabi, A., & Katsaggelos, A. K. (2019). Optical Flow Prediction for Blind and Non-Blind Video Error Concealment Using Deep Neural Networks. International Journal of Multimedia Data Engineering and Management (IJMDEM), 10(3), 27-46. http://doi.org/10.4018/IJMDEM.2019070102

Chicago

Sankisa, Arun, Arjun Punjabi, and Aggelos K. Katsaggelos. "Optical Flow Prediction for Blind and Non-Blind Video Error Concealment Using Deep Neural Networks," International Journal of Multimedia Data Engineering and Management (IJMDEM) 10, no.3: 27-46. http://doi.org/10.4018/IJMDEM.2019070102

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

A novel optical flow prediction model using an adaptable deep neural network architecture for blind and non-blind error concealment of videos degraded by transmission loss is presented. The two-stream network model is trained by separating the horizontal and vertical motion fields which are passed through two similar parallel pipelines that include traditional convolutional (Conv) and convolutional long short-term memory (ConvLSTM) layers. The ConvLSTM layers extract temporally correlated motion information while the Conv layers correlate motion spatially. The optical flows used as input to the two-pipeline prediction network are obtained through a flow generation network that can be easily interchanged, increasing the adaptability of the overall end-to-end architecture. The performance of the proposed model is evaluated using real-world packet loss scenarios. Standard video quality metrics are used to compare frames reconstructed using predicted optical flows with those reconstructed using “ground-truth” flows obtained directly from the generator.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.