This paper provides an overview of the state-of-the-art in laboratory speaker-independent, large vocabulary continuous speech recognition (LVCSR) systems with a view towards adapting such technology to the requirements of real-world applications. While in speech recognition the principal concern is to transcribe the speech signal as a sequence of words, the same core technology can be applied to domains other than dictation. The main topics addressed are acoustic-phonetic modeling, lexical representation, language modeling, decoding and model adaptation. After a brief summary of experimental results some directions towards usable systems are given. In moving from laboratory systems towards real-world applications, different constraints arise which influence the system design. The application imposes limitations on computational resources, constraints on signal capture, requirements for noise and channel compensation, and rejection capability. The difficulties and costs of adapting existing technology to new languages and application need to be assessed. Near term applications for LVCSR technology are likely to grow in somewhat limited domains such as spoken language systems for information retrieval, and limited domain dictation. Perspectives on some unresolved problems are given, indicating areas for future research

In this paper we propose a machine-learning approach to paragraph boundary identification which utilizes linguistically motivated features. We investigate the relation between paragraph boundaries and discourse cues, pronominalization and information structure. We test our algorithm on German data and report improvements over three baselines including a reimplementation of Sporleder & Lapata’s (2006) work on paragraph segmentation. An analysis of the features’ contribution suggests an interpretation of what paragraph boundaries indicate and what they depend on.

This paper considers the problem of automatic paragraph segmentation. The task is relevant for speech-to-text applications whose output transcipts do not usually contain punctuation or paragraph indentation and are naturally difficult to read and process. Text-to-text generation applications (e.g., summarisation) could also benefit from an automatic paragaraph segementation mechanism which indicates topic shifts and provides visual targets to the reader. We present a paragraph segmentation model which exploits a variety of knowledge sources (including textual cues, syntactic and discourse related information) and evaluate its performance in different languages and domains. Our experiments demonstrate that the proposed approach significantly outperforms our baselines and in many cases comes to within a few percent of human performance. Finally, we integrate our method with a single document summariser and show that it is useful for structuring the output of automatically generated text.

This paper describes NYU's effort toward improving recognition accuracy for the 1996 ARPA Large Vocabulary Continuous Speech Recognition evaluation. We are trying to develop different kinds of language models including longer-range models and a linguistically motivated model. For the system described here, we used as a starting point the scores produced by SRI's acoustic and language models. These are linearly combined with the scores produced by the NYU language models. This paper also describes some experiments we tried which were not used in the official experiment, including experiments with perplexity minimization, MaximumEntropy modeling and parsing. 1. Introduction This paper describes NYU's effort toward improving recognition accuracy for the 1996 ARPA Large Vocabulary Continuous Speech Recognition evaluation. Our goal has been to study some longerrange language models and determine whether they can be a useful component of the language models used for speech recognition. We ...