AI helps you reading Science
AI generates interpretation videos
AI extracts and analyses the key points of the paper to generate videos automatically
AI parses the academic lineage of this thesis
AI extracts a summary of this paper
Affective Mtv Analysis Based On Arousal And Valence Features
2008 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, VOLS 1-4, pp.1369-+, (2008)
- MTV (Music TV) is an important favorite pastime to modern people. Especially in recent years, MTV can be played conveniently on mobile sets including cell phones and music players such as iPod and Zune.
- The increasing amounts of MTV and storage capacity of the digital sets have caused many problems: how to effectively organize, manage and retrieve the desired MTVs. It is true that the traditional MTV classifications based on Artist, Album and Title, could be solutions to this problem.
- It is true that the traditional MTV classifications based on Artist, Album and Title, could be solutions to this problem
- These methods have many limitations when people want to manage and retrieve MTVs with semantic and abstract concepts.
- Users will be able to classify MTVs into categories according to MTVs’ affective states, so that they can select their desired categories to enjoy
- MTV (Music TV) is an important favorite pastime to modern people
- Affective MTV content analysis which has little been researched before might provide potential solutions to these problems
- We present a framework for affective MTV analysis
- The comparisons between our selected features and those in related work verify that our features improve the performance by a significant number
- The authors' dataset of MTV is representative.
- Results of Affective Clustering First, the Valence-based clustering is carried out and four clusters are generated.
- The Arousal-based clustering is carried out on these four pregenerated clusters respectively.
- Eight categories, within which the MTVs are similar in both Arousal and Valence, are generated.
- Each category is marked in the A-V space and colored, so user can identify each category’s affective state and select their desired categories to enjoy from MTV database
- The authors present a framework for affective MTV analysis.
- Six Arousal features and five Valence features are extracted.
- Affinity Propagation is utilized to put MTVs with similar affective states into same categories.
- The authors conduct subjective user study to obtain the background truth about the affective state of each MTV.
- The numerical evaluations prove the validity of the framework.
- The comparisons between the selected features and those in related work verify that the features improve the performance by a significant number
- Table1: The options and corresponding descriptions
- Table2: Rules for ground truth computation
- Table3: The quantified affective states of each cluster
- Table4: The numerical evaluation results
- Table5: The comparison of A-V features
- Table6: The precision improvement
- The authors gratefully acknowledge the support of K
- Wong Education Foundation, Hong Kong and Science100 Plan of Chinese Academy of Sciences under Grant 99T3002T03
- . A. Hanjalic, “Extracting Moods from Pictures and Sounds: Towards Truly Personalized TV,” IEEE Signal Processing Magazine, pp. 90 - 100, Mar. 2006.
- . A. Hanjalic and L.Q. Xu, “Affective Video Content Representation and Modeling,” IEEE Transactions on Multimedia, pp.143 – 154, Feb. 2005.
- . A. Hanjalic, “Adaptive Extraction of Highlights from a Sport Video Based on Excitement Modeling,” IEEE Transactions on Multimedia, pp.1114 – 1122, Dec. 2005.
- . M. Bradley, “Emotional Memory: A Dimensional Analysis,” Essays on Emotional Theory, Hillsdale, NJ: LEA, pp. 97–134, 1994.
- . J. Russell and A. Mehrabian, Evidence for a Three-Factor Theory of Emotions, J. Res. Person, vol. 11:273–294, 1977.
- . H. Schlosberg, “Three Dimensions of Emotion,” Psychol. Rev, vol. 61, no. 2, pp. 81–88, Mar. 1954.
- . L. Lu, D. Liu and H.J. Zhang, “Automatic Mood Detection and Tracking of Music Audio Signals,” IEEE Transactions on Audio and Language Processing, Vol. 14(1), January, 2006.
- . B. J. Frey and D. Dueck, “Clustering by Passing Messages Between Data Points,” SECINCE, Vol.315:972-976,Feb. 2007.
- . T. Zhang and C.-C. Jay Kuo, “Audio Content Analysis for Online Audiovisual Data Segmentation and Classification,” IEEE Transactions on Speech And Audio Processing, Vol. 9(4), pp. 441-457, MAY, 2001.
- . N. Maddage, C. S. Xu, M.S. Kankanhalli and X. Shao, “Content-based Music Structure Analysis with Applications to Music Semantics Understanding,” Proc. of ACM Multimedia, pp. 112-119, Oct. 2004.
- . H. L. Wang and L.F. Cheong, “Affective Understanding in Film,” IEEE Transactions on Circuits and Systems for Video Technology, Vol. 16(6), pp. 689-704, June, 2006.
- . D. Bordwell and K. Thompson, Film Art: An Introduction, 7rd ed. New York: McGraw-Hill, 2004.
- . H. Zettl, Sight Sound Motion: Applied Media Aesthetics, 3rd ed. Belmont, CA: Wadsworth, 1998.