Model-based Sequential Organization in Cochannel Speech
Abstract
A human listener has the ability to follow a speakers voice while others are speaking simultaneously; in particular, the listener can organize the time-frequency energy of the same speaker across time into a single stream. In this paper, we focus on sequential organization in cochannel speech, or mixtures of two voices. We extract minimally corrupted segments, or usable speech, in cochannel speech using a robust multipitch tracking algorithm. The extracted usable speech is shown to capture speaker characteristics and improves speaker identification performance across various target-to-interferer ratios. To utilize speaker characteristics for sequential organization, we extend the traditional speaker identification framework to cochannel speech and derive a joint objective for sequential grouping and speaker identification, leading to a problem of search for the optimum hypothesis. Subsequently we propose a hypothesis pruning algorithm based on speaker models in order to make the search computationally feasible. Evaluation results show that the proposed system approaches the ceiling speaker identification performance obtained with prior pitch information, and yields significant improvement over alternative approaches on sequential organization.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 2004
- Accession Number
- AD1001181
Entities
People
- DeLiang Wang
- Yang Shao
Organizations
- Ohio State University