Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning

There is an abundance of digital video content due to the cloud’s phenomenal growth and security footage, it is therefore essential to summarize these videos in data centers. This paper offers innovative approaches to the problem of key-frame extraction for the purpose of video summarization. Our ap...

Full description

Saved in:
Bibliographic Details
Main Author: Issa, Obada (author)
Other Authors: Shanableh, Tamer (author)
Format: article
Published: 2023
Subjects:
Online Access:http://hdl.handle.net/11073/25249
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1864513432890376192
author Issa, Obada
author2 Shanableh, Tamer
author2_role author
author_facet Issa, Obada
Shanableh, Tamer
author_role author
dc.creator.none.fl_str_mv Issa, Obada
Shanableh, Tamer
dc.date.none.fl_str_mv 2023-05-15T09:55:30Z
2023-05-15T09:55:30Z
2023
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv Issa O, Shanableh T. Static Video Summarization Using Video Coding Features with Frame-Level Temporal Subsampling and Deep Learning. Applied Sciences. 2023; 13(10):6065.
2076-3417
http://hdl.handle.net/11073/25249
10.3390/app13106065
dc.language.none.fl_str_mv en_US
dc.publisher.none.fl_str_mv MDPI
dc.relation.none.fl_str_mv https://doi.org/10.3390/app13106065
dc.subject.none.fl_str_mv Video summarization
Video coding
Temporal subsampling
Convolution neural networks
Long-short term memory
dc.title.none.fl_str_mv Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
dc.type.none.fl_str_mv Peer-Reviewed
Postprint
info:eu-repo/semantics/publishedVersion
info:eu-repo/semantics/article
description There is an abundance of digital video content due to the cloud’s phenomenal growth and security footage, it is therefore essential to summarize these videos in data centers. This paper offers innovative approaches to the problem of key-frame extraction for the purpose of video summarization. Our approach includes feature variables extracted from the bit streams of coded videos, followed by optional stepwise regression for dimensionality reduction. Once the features are extracted and reduced in dimensionality, we apply innovate frame-level temporal sub-sampling techniques followed by training and testing using deep learning architectures. The frame-level temporal subsampling techniques are based on cosine similarity and PCA projections of feature vectors. We create three different learning architectures by utilizing LSTM networks, 1D-CNN networks, and Random Forests. The four most popular video summarization datasets, namely, TVSum, SumMe, OVP, and VSUMM are used to evaluate the accuracy of the proposed solutions. This includes the Precision, Recall, F-score measures, and computational time. It is shown that the proposed solutions when trained and tested on all subjective user summaries, achieved F-scores of 0.79, 0.74, 0.88, and 0.81, respectively, for the aforementioned datasets, showing clear improvements over prior studies.
format article
id aus_f8c14f1a3a63b53761fd9fc461288bfd
identifier_str_mv Issa O, Shanableh T. Static Video Summarization Using Video Coding Features with Frame-Level Temporal Subsampling and Deep Learning. Applied Sciences. 2023; 13(10):6065.
2076-3417
10.3390/app13106065
language_invalid_str_mv en_US
network_acronym_str aus
network_name_str aus
oai_identifier_str oai:repository.aus.edu:11073/25249
publishDate 2023
publisher.none.fl_str_mv MDPI
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
spelling Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep LearningIssa, ObadaShanableh, TamerVideo summarizationVideo codingTemporal subsamplingConvolution neural networksLong-short term memoryThere is an abundance of digital video content due to the cloud’s phenomenal growth and security footage, it is therefore essential to summarize these videos in data centers. This paper offers innovative approaches to the problem of key-frame extraction for the purpose of video summarization. Our approach includes feature variables extracted from the bit streams of coded videos, followed by optional stepwise regression for dimensionality reduction. Once the features are extracted and reduced in dimensionality, we apply innovate frame-level temporal sub-sampling techniques followed by training and testing using deep learning architectures. The frame-level temporal subsampling techniques are based on cosine similarity and PCA projections of feature vectors. We create three different learning architectures by utilizing LSTM networks, 1D-CNN networks, and Random Forests. The four most popular video summarization datasets, namely, TVSum, SumMe, OVP, and VSUMM are used to evaluate the accuracy of the proposed solutions. This includes the Precision, Recall, F-score measures, and computational time. It is shown that the proposed solutions when trained and tested on all subjective user summaries, achieved F-scores of 0.79, 0.74, 0.88, and 0.81, respectively, for the aforementioned datasets, showing clear improvements over prior studies.American University of SharjahMDPI2023-05-15T09:55:30Z2023-05-15T09:55:30Z2023Peer-ReviewedPostprintinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfIssa O, Shanableh T. Static Video Summarization Using Video Coding Features with Frame-Level Temporal Subsampling and Deep Learning. Applied Sciences. 2023; 13(10):6065.2076-3417http://hdl.handle.net/11073/2524910.3390/app13106065en_UShttps://doi.org/10.3390/app13106065oai:repository.aus.edu:11073/252492024-08-22T12:07:18Z
spellingShingle Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
Issa, Obada
Video summarization
Video coding
Temporal subsampling
Convolution neural networks
Long-short term memory
status_str publishedVersion
title Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
title_full Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
title_fullStr Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
title_full_unstemmed Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
title_short Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
title_sort Static Video Summarization Using Video Coding Features with Frame-level Temporal Sub-Sampling and Deep Learning
topic Video summarization
Video coding
Temporal subsampling
Convolution neural networks
Long-short term memory
url http://hdl.handle.net/11073/25249