Author: SRI International

We present an N-best rescoring algorithm that removes the effect of segmentation mismatch. Furthermore, we show that explicit language modeling of hidden linguistic segment boundaries is improved by including turn-boundary events in the model.

In the context of automatic speaker recognition, we propose a model transformation technique that renders speaker models more robust to acoustic mismatches and to data scarcity by appropriately increasing their variances.

This paper studies the effects of handset distortion on telephone-based speaker recognition performance. Results on the 1996 NIST Speaker Recognition Evaluation corpus show that using handset-matched background models reduces false acceptances (at a 10% miss rate) by more than 60% over previously reported (handset-independent) approaches.

This paper proposes a probabilstic framework to define and evaluate confidence measures for word recognition. We describe a novel method to combine different knowledge sources and estimate the confidence in a word hypothesis, via a neural network.

This paper presents a speech-enabled WWW demonstration based on the Air Travel Information System (ATIS) domain. SRI’s speech recognition technology and natural language understanding are fully integrated in a Java application using the DECIPHER(TM) speech recognition system and the Open Agent Architecture(TM).

We describe the development of the SRI system evaluated in the 1996 DARPA continuous speech recognition (CSR) Hub4 partitioned evaluation (PE). The task for the Hub4 evaluation was to recognition speech from broadcast television and radio shows.

In SRI’s language modeling experiments for the Hub4 domain, three basic approaches were pursued: interpolating multiple models estimated from Hub4 and non-Hub4 training data, adapting the language model (LM) to the focus conditions, and adapting the LM to different topic types.