System and method for automatic audio content analysis for word spotting, indexing, classification and retrieval

摘要: A system and method for indexing an audio stream subsequent information retrieval skimming, gisting, summarizing the includes using special prefiltering such that only relevant speech segments are generated by a recognition engine indexed. Specific features disclosed improve precision recall of used after word spotting. The invention rendering into intervals, with each interval including one or more segments. For segment it is determined whether exhibits predetermined as particular range zero crossing rates, energy, spectral energy concentration. heuristically to represent respective events silence, music, speech, on music. Also, group intervals matches predefined meta pattern continuous uninterrupted concluding ideas, hesitations emphasis in so on, then indexed based classification matching, being retrieval. alternatives longer terms along weights, recall.