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.

Low-temperature growth of carbon nanotubes by plasma-enhanced chemical vapor deposition

Vertically aligned carbon nanotubes were grown at temperatures as low as 120degreesC by plasma-enhanced chemical vapor deposition. A systematic study of the temperature dependence of the growth rate and the structure of the as-grown nanotubes is presented using a C2H2/NH3 system and nickel as the catalyst. The activation energy for the growth rate was found to be 0.23 eV, much less than for thermal chemical vapor deposition (1.2-1.5 eV). This suggests growth occurs by surface diffusion of carbon on nickel. The result could allow direct growth of nanotubes onto low-temperature substrates like plastics, and facilitate the integration in sensitive nanoelectronic devices. (C) 2003 American Institute of Physics.