2008
DOI: 10.1109/tsmcc.2008.919173
|View full text |Cite
|
Sign up to set email alerts
|

Automatic Video Classification: A Survey of the Literature

Abstract: Abstract-There is much video available today. To help viewers find video of interest, work has begun on methods of automatic video classification. In this paper, we survey the video classification literature. We find that features are drawn from three modalities-text, audio, and visual-and that a large variety of combinations of features and classification have been explored. We describe the general features chosen and summarize the research in this area. We conclude with ideas for further research.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
157
1
12

Year Published

2012
2012
2022
2022

Publication Types

Select...
6
3
1

Relationship

0
10

Authors

Journals

citations
Cited by 243 publications
(170 citation statements)
references
References 68 publications
0
157
1
12
Order By: Relevance
“…In [22] the authors address the problem of video genres classification for the five classes with a set of visual features, and SVM is used for classification. Huge literature reports can be obtained for automatic video classification in [4]. Several audio-visual features have been described for characterzing semantic content in multimedia in [25].The edge based feature, namely, the percentage of edge pixels, is extracted from each key frame for classifying a given sports video into one of the five categories, namely, badmiton, soccer, basket ball, tennis and figure skating techniques as explained in [30].…”
Section: Related Workmentioning
confidence: 99%
“…In [22] the authors address the problem of video genres classification for the five classes with a set of visual features, and SVM is used for classification. Huge literature reports can be obtained for automatic video classification in [4]. Several audio-visual features have been described for characterzing semantic content in multimedia in [25].The edge based feature, namely, the percentage of edge pixels, is extracted from each key frame for classifying a given sports video into one of the five categories, namely, badmiton, soccer, basket ball, tennis and figure skating techniques as explained in [30].…”
Section: Related Workmentioning
confidence: 99%
“…Among these, some recent studies that analyze and compare the most relevant strategies can be found [11] [12] [13]. First proposals were focused on the detection of abrupt transitions (cuts) but, as these were located more efficiently, posterior strategies began to consider the detection of gradual transitions (dissolves, fades and wipes), which identification is more complex and difficult due to the many existing types of gradual transitions [14].…”
Section: Related Workmentioning
confidence: 99%
“…A state-of-the art is available in [1]. [8] discusses an uni-modal (image) approach and testes the prospective potential of motion information to cartoon classification.…”
Section: Introductionmentioning
confidence: 99%