Video annotation framework for news headlines using deep visual and text information

Luqman Izhar; Shazia Arshad; Urooj Saeed; Muhammad Usman Ghani Khan

doi:10.1109/iccubea.2017.8463790

Author(s)

Luqman Izhar Department, of Computer Science, UET, Lahore, Pakistan

Shazia Arshad Department, of Computer Science, UET, Lahore, Pakistan

Urooj Saeed Department, of Computer Science, UET, Lahore, Pakistan

Muhammad Usman Ghani Khan Department, of Computer Science, UET, Lahore, Pakistan

Abstract

TV News channels are broadcasting news bulletins every hour. News bulletins data set is growing each hour for video tagging and retrieval. News headlines bulletins are a combination of different news annotations. International news channels media houses rely on local news sources to track a news of a specific region. News team research group digs in manually to monitor trends of local news sources to determine if a news of a specific region can create a wave as international news. To select a specific news annotation from local news sources requires consideration of multiple real time factors like coverage, rating and impact. Proposed Research model solves this real time problem by using deep learning method .Framework proposed classifies and retrieves a video annotation based on news indexing, priority and quality. Data set of 250 news bulletins was gathered from top seven news channels for top ten global categories like politics, terrorism, finance, weather & entertainment etc. Framework classifies the news annotation retrieval and tagging using 2D CNN with validation accuracy of 98.14% by calculating and considering real time parameters news indexing, priority and quality across cross bulletins contributing to novelty of the research.

Publication Details

Page(s) 683-691

DOI 10.1109/iccubea.2017.8463790

Published Journal: Journal of Natural & Applied Sciences Pakistan, Volume: 3, Issue: 1, Year: 2021

Request the full text from PASTIC

Keywords

Indexing, Priority, Video, Headlines, 2DCNN, News, Annotation

References

Features Content . .. 2018 13th World Congress on Intelligent Control and Automation (WCICA)., : .

Roman L. S.,Monteiro N. T.,A. N. T. .2018 .. doi:10, : .

M. ,Rodan F.,Javed A.,Y. A. .2018 .Forecasting stock market index daily direction: A Bayesian Network approach. Expert Systems with Applications, 105 : 11-22.

2018 Fifth HCT Information Technology Trends (ITT). . .. , : .

Isah D.,Zulkernine H.,F. H. .2018 .. doi:10, : .

. .. 2018 IEEE International Conference on Big Data (Big Data)., : .

H. V. .2018 .Predicting the Popularity of Online News using Social Features. doi:10, : .

2018 Second International . .Conference on Green Computing and Internet of Things (ICGCIoT).. , : .