Unsupervised training for mandarin broadcast news and conversation transcription

A significant cost in obtaining acoustic training data is the generation of accurate transcriptions. For some sources close-caption data is available. This allows the use of lightly-supervised training techniques. However, for some sources and languages close-caption is not available. In these cases unsupervised training techniques must be used. This paper examines the use of unsupervised techniques for discriminative training. In unsupervised training automatic transcriptions from a recognition system are used for training. As these transcriptions may be errorful data selection may be useful. Two forms of selection are described, one to remove non-target language shows, the other to remove segments with low confidence. Experiments were carried out on a Mandarin transcriptions task. Two types of test data were considered, Broadcast News (BN) and Broadcast Conversations (BC). Results show that the gains from unsupervised discriminative training are highly dependent on the accuracy of the automatic transcriptions. © 2007 IEEE.

Thin film monitoring with silicon bulk acoustic resonators

Abstract-This paper reports a single-crystal silicon mass sensor based on a square-plate resonant structure excited in the wine glass bulk acoustic mode at a resonant frequency of 2.065 MHz and an impressive quality factor of 4 million at 12 mtorr pressure. Mass loading on the resonator results in a linear downshift in the resonant frequency of this device, wherein the measured sensitivity is found to be 175 Hz cm2/μg. The silicon resonator is embedded in an oscillator feedback loop, which has a short-term frequency stability of 3 mHz (approximately 1.5 ppb) at an operating pressure of 3.2 mtorr, corresponding to an equivalent mass noise floor of 17 pg/cm2. Possible applications of this device include thin film monitoring and gas sensing, with the potential added benefits of scalability and integration with CMOS technology. © 2008 IEEE.