Color-independent classification of animation video
Tóm tắt
This paper presents a method for the classification of animated video that does not rely on hue or saturation information, and aims to achieve a high level of performance in the context of automatic video categorization and encoder parameterization for low bit rate video processing. While existing animated and genre-based video classification approaches have achieved good results, they are highly dependent on color extraction and thus cannot be used to accurately analyze grayscale video content and realistic animation styles, such as pixilation, where the color palette is not sufficiently distinct from live action content. We first introduce a dataset specifically designed to test animation video classification with a high variety of visual content. We then present a set of hue-independent features based on spatial and temporal characteristics of animation video, which are used to perform the classification task. Most of these features have not previously been utilized in existing animation video classification approaches. Finally, an overview of test results, representing a success rate of over 80%, proves the viability of the proposed classification method.
Tài liệu tham khảo
Dammak N, Ben Ayed Y (2014) Video genre categorization using support vector machines. In: IEEE 1st international conference on advanced technologies for signal and image processing (ATSIP), vol. 1, pp. 106–110
Doulaty M, Saz O, W M Ng R, Hain T (2016) Automatic genre and show identification of broadcast media. In: Proceedings of the 17th annual conference of the international speech communication association (Interspeech), vol. 1, pp. I–I
Duarte LA, Penatti OAB, Almeida J (2016) Bag of genres for video retrieval. In: IEEE 29th SIBGRAPI conference on graphics, patterns and images (SIBGRAPI), vol. 1, pp. 257–264
Ekenel HK, Semela T, Stiefelhagen R (2010) Content-based video genre classification using multiple cues. In: Proceedings of the 3rd international workshop on Automated information extraction in media production ACM, pp. 21–26
Hamed AAM, Li R, Xiaoming Z, Xu C (2013) Video genre classification using weighted kernel logistic regression. Advances in Multimedia, pp. I–I
Ianeva TI, de Vries AP, Rohrig H (2003) Detecting cartoons: a case study in automatic video-genre classification. In: IEEE proceedings 2003 international conference on multimedia and expo, 2003. ICME’03, vol. 1, pp. I–449
Ionescu B, Lambert P (2013) Classification of animated video genre using color and temporal information. UPB Sci Bull Ser C 75(3):63–74
Ionescu B, Lambert P, Coquin D, Buzuloiu V (2007) The cut detection issue in the animation movie domain. J Multimed, 2(4)
Ionescu B, Seyerlehner K, Rasche C, Vertan C, Lambert P (2012) Content-based video description for automatic video genre categorization. Adv Multimed Model, pp. 51–62
Ionescu BE, Seyerlehner K, Mironic I, Vertan C, Lambert P (2014) An audio-visual approach to web video categorization. Multimed Tools Appl, pp. 1007–1032
Jang WD, Lee C, Sim JY, Kim CS (2014) Automatic video genre classification using multiple svm votes. In: IEEE 22nd international conference on pattern recognition (ICPR), 2014 , vol. 1, pp. 2655–2660
Kim S, Panayiotis G, Narayanan S (2013) On-line genre classification of tv programs using audio content. In: IEEE international conference on acoustics, speech and signal processing (ICASSP) 2013, vol. 1, pp. 798–802
Lienhart R, Maydt J (2002) An extended set of haar-like features for rapid object detection. In: IEEE proceedings international conference on image processing, vol. 1, pp. I–I
Mironic I, Du IC, Ionescu B, Sebe N (2016) A modified vector of locally aggregated descriptors approach for fast video classification. Multimed Tools Appl, pp. 9045–9072
Rouvier M, Oger S, Linars G, Matrouf D, Merialdo B, Li Y (2015) Audio-based video genre identification. IEEE ACM Trans Audio Speech Language Process, pp. 1031–1041
Safdarnejad SM, Liu X, Udpa L (2014) Genre categorization of amateur sports videos in the wild. In: IEEE international conference on image processing (ICIP) vol. 1, pp. 1001–1005
Saz O, Doulaty M, Hain T (2014) Background-tracking acoustic features for genre identification of broadcast shows. In: IEEE spoken language technology workshop (SLT), vol. 1, pp. 118–123
Simes GS, Wehrmann J, Barros RC, Ruiz DD (2016) Movie genre classification with convolutional neural networks. In: IEEE international joint conference on neural networks (IJCNN), vol. 1, pp. 259–266
Uijlings J, Duta IC, Sangineto E, Sebe N (2015) Video classification with densely extracted hog/hof/mbh features: an evaluation of the accuracy/computational efficiency trade-off. Int J Multimed Information Retr, pp. 33–44
Viola P, Jones M (2001) Rapid object detection using a boosted cascade of simple features. In: Proceedings of the 2001 IEEE computer society conference on computer vision and pattern recognition CVPR 2001, vol. 1, pp. I–I
Wehrmann J, Barros RC (2017) Convolutions through time for multi-label movie genre classification. In: SAC ’17 proceedings of the symposium on applied computing, vol. 1, pp. 114–119
Wehrmann J, Barros RC, Simes GS, Paula TS, Ruiz DD (2016) (deep) learning from frames. In: 5th Brazilian conference on intelligent systems (BRACIS), vol. 1, pp. 1–6
Yuan X, Lai W, Mei T, Hua XS, Wu XQ, Li S (2006) Automatic video genre categorization using hierarchical svm. In: IEEE international conference on image processing, pp. 2905–2908