Our objectives in the DARPA Spoken Language Program are: (1) to provide a central role in speech database design, development and distribution; and (2) to design, coordinate implementation of, and analyze results of performance tests. (3) Prepared the Prototype CD-ROM Version of the DARPA TIMIT Acoustic-Phonetic Speech Database. This CD-ROM contains the data for 420 talkers (4200 sentence utter...

With the demands of globalization, multilingual speech is increasingly common in conversational telephone speech, broadcast news and internet podcasts. Therefore, automatic spoken language recognition has become an important technology in multilingual speech related applications. For example, automatic spoken language recognition has been used as a preprocessing component for spoken language tr...

In the future network era, huge volume of information on all subject domains will be readily available via the network. Also, all the network information are dynamic, ever-changing and exploding. Furthermore, many of the spoken language processing applications will have to do with the content of the network information, which is dynamic. This means dynamic lexicons, language models and so on wi...

This study investigates age related differences in standardized tests scores of language usage and reading from elementary to high school for students who are either monolinguals whose L1 is English or bilinguals whose L1 is not English. An interactioneffect between grade level andreading and language usage standardized test scores was hypothesized because as bilinguals become proficient in Cog...

In the speech-to-speech translation framework automatic speech recognition and spoken language translation components provide additional information about the location of pauses in the source language. This information may be useful to improve the performance of pause prediction algorithms for speech synthesis. In this paper we propose a transfer algorithm based on tuples. The results show a be...

In this paper. we argue for a paradigm shift in spoken language technology, from transcription tasks to interactive systems. The current paradigm evaluates speech recognition technology in terms of word recognition accuracy on large vocabulary transcription tasks? such as telephone conversations or media broadcasts. Systems are evaluated in international competitions, with strict rules for part...

This paper discuss a dependency analyzer employing Support Vector Machines (SVMs) for Japanese spoken language. Most conventional dependency analyzers target written texts. Thus, we use a currently available spoken language corpus and make the SVMs learn the corpus to build a dependency analyzer that targets spoken language. We used two types of corpora: one contains written language, and the o...

The i-vector technique using deep neural network has been successfully applied in spoken language identification systems. Neural network modeling showed its effectiveness as both discriminant feature transformation and classification in many tasks, in particular with a large training data set. However, on a small data set, neural networks suffer from the overfitting problem which degrades the p...

In human-mediated translation scenarios a human interpreter translates between a source and a target language using either a spoken or a written representation of the source language. In this paper we improve the recognition performance on the speech of the human translator spoken in the target language by taking advantage of the source language representations. We use machine translation techn...