Nobuyasu Itoh

Learn More
This paper presents a strategy for efficiently selecting informative data from large corpora of untranscribed speech. Confidence-based selection methods (i.e., selecting utterances we are least confident about) have been a popular approach, though they only look at the top hypothesis when selecting utterances and tend to select outliers, therefore, not(More)
This paper introduces a discriminative training for language models (LMs) by leveraging phoneme similarities estimated from an acoustic model. To train an LM discriminatively, we needed the correct word sequences and the recognized results that Automatic Speech Recognition (ASR) produced by processing the utterances of those correct word sequences. But,(More)
-This paper describes a method of spelling correction consisting of two steps: selection of candidate words, and approximate string matching between the input word and each candidate word. Each word is classified and multi-indexed according to combinations of a constant number of characters in the word. Candidate words are selected fast and accurately,(More)
The composition of the dopant for the analysis of polycyclic aromatic hydrocarbons (PAHs) by liquid chromatography/dopant-assisted atmospheric-pressure photoionization/mass spectrometry under reversed-phase conditions was optimized to enhance the ionization efficiency for PAHs. The most suitable dopant was a toluene/anisole mixture (99.5:0.5, v/v) and it(More)
For accurate quantification of polycyclic aromatic hydrocarbons (PAHs) in dust samples, we investigated the use of microwave-assisted solvent extraction (MAE) combined with isotope-dilution mass spectrometry (IDMS) using deuterium-labelled PAHs (D-PAHs). Although MAE with a methanol/toluene mixture (1:3 by volume) at 160°C for 40 min was best for extracting(More)
In this paper we present an extention of a context tree for a structured language model (SLM), which we call an arbori-context tree. The state-of-the-art SLM predicts the next word from a xed partial tree of the history tree, such as two exposed heads, etc. An arbori-context tree allows us to select an optimum partial tree of a history tree for the next(More)
Document recognition system (DRS), a workstation-based prototype document analysis system that uses optical character recognition (OCR), is described. The system provides functions for image capture, block segmentation, page structure analysis, and character recognition with contextual postprocessing, as well as a user interface for error correction. All(More)
Microwave-assisted extraction using 1M KOH/methanol (alkaline-MAE) in combination with solid-phase extraction treatment was developed and applied to polycyclic aromatic hydrocarbons (PAHs) in a sediment sample. Although various conditions were examined (100 or 150 degrees C for 10 or 30 min), comparable concentrations of PAHs to those obtained by(More)