Activity-driven content adaptation for effective video summarisation

Jinchang Ren, J. Jiang, Y. Feng

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

In this paper, we present a novel method for content adaptation and video summarization fully implemented in compressed-domain. Firstly, summarization of generic videos is modeled as the process of extracted human objects under various activities/events. Accordingly, frames are classified into five categories via fuzzy decision including shot changes (cut and gradual transitions), motion activities (camera motion and object motion) and others by using two inter-frame measurements. Secondly, human objects are detected using Haar-like features. With the detected human objects and attained frame categories, activity levels for each frame are determined to adapt with video contents. Continuous frames belonging to same category are grouped to form one activity entry as content of interest (COI) which will convert the original video into a series of activities. An overall adjustable quota is used to control the size of generated summarization for efficient streaming purpose. Upon this quota, the frames selected for summarization are determined by evenly sampling the accumulated activity levels for content adaptation. Quantitative evaluations have proved the effectiveness and efficiency of our proposed approach, which provides a more flexible and general solution for this topic as domain-specific tasks such as accurate recognition of objects can be avoided.
LanguageEnglish
Pages930-938
Number of pages9
JournalJournal of Visual Communication and Image Representation
Volume21
Issue number8
DOIs
Publication statusPublished - Nov 2010

Fingerprint

Cameras
Sampling

Keywords

  • image representation
  • content adaptation
  • video summarization

Cite this

@article{8cee55fd436549dea83241b53a6c8593,
title = "Activity-driven content adaptation for effective video summarisation",
abstract = "In this paper, we present a novel method for content adaptation and video summarization fully implemented in compressed-domain. Firstly, summarization of generic videos is modeled as the process of extracted human objects under various activities/events. Accordingly, frames are classified into five categories via fuzzy decision including shot changes (cut and gradual transitions), motion activities (camera motion and object motion) and others by using two inter-frame measurements. Secondly, human objects are detected using Haar-like features. With the detected human objects and attained frame categories, activity levels for each frame are determined to adapt with video contents. Continuous frames belonging to same category are grouped to form one activity entry as content of interest (COI) which will convert the original video into a series of activities. An overall adjustable quota is used to control the size of generated summarization for efficient streaming purpose. Upon this quota, the frames selected for summarization are determined by evenly sampling the accumulated activity levels for content adaptation. Quantitative evaluations have proved the effectiveness and efficiency of our proposed approach, which provides a more flexible and general solution for this topic as domain-specific tasks such as accurate recognition of objects can be avoided.",
keywords = "image representation, content adaptation, video summarization",
author = "Jinchang Ren and J. Jiang and Y. Feng",
year = "2010",
month = "11",
doi = "10.1016/j.jvcir.2010.09.002",
language = "English",
volume = "21",
pages = "930--938",
journal = "Journal of Visual Communication and Image Representation",
issn = "1047-3203",
number = "8",

}

Activity-driven content adaptation for effective video summarisation. / Ren, Jinchang; Jiang, J.; Feng, Y.

In: Journal of Visual Communication and Image Representation, Vol. 21, No. 8, 11.2010, p. 930-938.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Activity-driven content adaptation for effective video summarisation

AU - Ren, Jinchang

AU - Jiang, J.

AU - Feng, Y.

PY - 2010/11

Y1 - 2010/11

N2 - In this paper, we present a novel method for content adaptation and video summarization fully implemented in compressed-domain. Firstly, summarization of generic videos is modeled as the process of extracted human objects under various activities/events. Accordingly, frames are classified into five categories via fuzzy decision including shot changes (cut and gradual transitions), motion activities (camera motion and object motion) and others by using two inter-frame measurements. Secondly, human objects are detected using Haar-like features. With the detected human objects and attained frame categories, activity levels for each frame are determined to adapt with video contents. Continuous frames belonging to same category are grouped to form one activity entry as content of interest (COI) which will convert the original video into a series of activities. An overall adjustable quota is used to control the size of generated summarization for efficient streaming purpose. Upon this quota, the frames selected for summarization are determined by evenly sampling the accumulated activity levels for content adaptation. Quantitative evaluations have proved the effectiveness and efficiency of our proposed approach, which provides a more flexible and general solution for this topic as domain-specific tasks such as accurate recognition of objects can be avoided.

AB - In this paper, we present a novel method for content adaptation and video summarization fully implemented in compressed-domain. Firstly, summarization of generic videos is modeled as the process of extracted human objects under various activities/events. Accordingly, frames are classified into five categories via fuzzy decision including shot changes (cut and gradual transitions), motion activities (camera motion and object motion) and others by using two inter-frame measurements. Secondly, human objects are detected using Haar-like features. With the detected human objects and attained frame categories, activity levels for each frame are determined to adapt with video contents. Continuous frames belonging to same category are grouped to form one activity entry as content of interest (COI) which will convert the original video into a series of activities. An overall adjustable quota is used to control the size of generated summarization for efficient streaming purpose. Upon this quota, the frames selected for summarization are determined by evenly sampling the accumulated activity levels for content adaptation. Quantitative evaluations have proved the effectiveness and efficiency of our proposed approach, which provides a more flexible and general solution for this topic as domain-specific tasks such as accurate recognition of objects can be avoided.

KW - image representation

KW - content adaptation

KW - video summarization

UR - http://www.journals.elsevier.com/journal-of-visual-communication-and-image-representation

U2 - 10.1016/j.jvcir.2010.09.002

DO - 10.1016/j.jvcir.2010.09.002

M3 - Article

VL - 21

SP - 930

EP - 938

JO - Journal of Visual Communication and Image Representation

T2 - Journal of Visual Communication and Image Representation

JF - Journal of Visual Communication and Image Representation

SN - 1047-3203

IS - 8

ER -