Title :
Significance of anchor speaker segments for constructing extractive audio summaries of broadcast news
Author :
Yella, Sree Harsha ; Varma, Vasudeva ; Prahallad, Kishore
Author_Institution :
Int. Inst. of Inf. Technol., Hyderabad, India
Abstract :
Analysis of human reference summaries of broadcast news showed that humans give preference to anchor speaker segments while constructing a summary. Therefore, we exploit the role of anchor speaker in a news show by tracking his/her speech to construct indicative/informative extractive audio summaries. Speaker tracking is done by Bayesian information criterion (BIC) technique. The proposed technique does not require Automatic Speech Recognition (ASR) transcripts or human reference summaries for training. The objective evaluation by ROUGE showed that summaries generated by the proposed technique are as good as summaries generated by a baseline text summarization system taking manual transcripts as input and summaries generated by a supervised speech summarization system trained using human summaries. The subjective evaluation of audio summaries by humans showed that they prefer summaries generated by proposed technique to summaries generated by supervised speech summarization system.
Keywords :
Bayes methods; multimedia communication; speech recognition; Bayesian information criterion technique; anchor speaker segment; automatic speech recognition transcript; baseline text summarization system; broadcast news; human reference summaries; informative extractive audio summaries; supervised speech summarization system; Broadcast news summarization; Speaker tracking; Summarization;
Conference_Titel :
Spoken Language Technology Workshop (SLT), 2010 IEEE
Conference_Location :
Berkeley, CA
Print_ISBN :
978-1-4244-7904-7
Electronic_ISBN :
978-1-4244-7902-3
DOI :
10.1109/SLT.2010.5700815
