论文信息 - Audio Segmentation System for Sport Games

Audio Segmentation System for Sport Games

This paper proposes a two-pass audio segmentation method for sports games. The 1st pass conducts the segmentation by a metric-based algorithm, and the 2nd pass conducts a model-based classification to extract speech segments. This audio segmentation module we developed can extract announcer's speech efficiently from the complex sport audio stream.

[1] Lie Lu,et al. Real-time unsupervised speaker change detection , 2002, Object recognition supported by user interaction for service robots.

[2] H Hermansky,et al. Perceptual linear predictive (PLP) analysis of speech. , 1990, The Journal of the Acoustical Society of America.

[3] Stan Z. Li,et al. Content-based Classification and Retrieval of Audio Using the Nearest Feature Line Method , 2000 .

[4] Yasuo Ariki,et al. Live speech recognition in sports games by adaptation of acoustic model and language model , 2003, INTERSPEECH.

[5] Douglas A. Reynolds,et al. Robust text-independent speaker identification using Gaussian mixture speaker models , 1995, IEEE Trans. Speech Audio Process..

[6] Jeih-Weih Hung,et al. Automatic metric-based speech segmentation for broadcast news via principal component analysis , 2000, INTERSPEECH.

[7] Stan Z. Li,et al. Content-based audio classification and retrieval using the nearest feature line method , 2000, IEEE Trans. Speech Audio Process..

[8] Lie Lu,et al. Digital Object Identifier (DOI) 10.1007/s00530-002-0065-0 Multimedia Systems , 2003 .