Unsupervised intra-lingual and cross-lingual speaker adaptation for HMM-based speech synthesis using two-pass decision tree construction

Hidden Markov model (HMM)-based speech synthesis systems possess several advantages over concatenative synthesis systems. One such advantage is the relative ease with which HMM-based systems are adapted to speakers not present in the training dataset. Speaker adaptation methods used in the field of HMM-based automatic speech recognition (ASR) are adopted for this task. In the case of unsupervised speaker adaptation, previous work has used a supplementary set of acoustic models to estimate the transcription of the adaptation data. This paper first presents an approach to the unsupervised speaker adaptation task for HMM-based speech synthesis models which avoids the need for such supplementary acoustic models. This is achieved by defining a mapping between HMM-based synthesis models and ASR-style models, via a two-pass decision tree construction process. Second, it is shown that this mapping also enables unsupervised adaptation of HMM-based speech synthesis models without the need to perform linguistic analysis of the estimated transcription of the adaptation data. Third, this paper demonstrates how this technique lends itself to the task of unsupervised cross-lingual adaptation of HMM-based speech synthesis models, and explains the advantages of such an approach. Finally, listener evaluations reveal that the proposed unsupervised adaptation methods deliver performance approaching that of supervised adaptation.

A condition monitoring platform using COTS wireless sensor networks - Lessons and experience

Developments in Micro-Electro-Mechanical Systems (MEMS), wireless communication systems and ad-hoc networking have created new dimensions to improve asset management not only during the operational phase but throughout an asset's lifecycle based on using improved quality of information obtained with respect to two key aspects of an asset: its location and condition. In this paper, we present our experience as well as lessons learnt from building a prototype condition monitoring platform to demonstrate and to evaluate the use of COTS wireless sensor networks to develop a prototype condition monitoring platform with the aim of improving asset management by providing accurate and real-time information. © 2010 IEEE.

Zinc oxide nanowire based hydrogen sensor on SOI CMOS platform

Here we report on the successful low-temperature growth of zinc oxide nanowires (ZnONWs) on silicon-on-insulator (SOI) CMOS micro-hotplates and their response, at different operating temperatures, to hydrogen in air. The SOI micro-hotplates were fabricated in a commercial CMOS foundry followed by a deep reactive ion etch (DRIE) in a MEMS foundry to form ultra-low power membranes. The micro-hotplates comprise p+ silicon micro-heaters and interdigitated metal electrodes (measuring the change in resistance of the gas sensitive nanomaterial). The ZnONWs were grown as a post-CMOS process onto the hotplates using a CMOS friendly hydrothermal method. The ZnONWs showed a good response to 500 to 5000 ppm of hydrogen in air. We believe that the integration of ZnONWs with a MEMS platform results in a low power, low cost, hydrogen sensor that would be suitable for handheld battery-operated gas sensors. © 2011 Published by Elsevier Ltd.