Abstract: A semiconductor device includes a combination substrate having a bulk silicon region, and a silicon-on-insulator (SOI) region. The SOI region includes a crystallized silicon layer formed by annealing amorphous silicon and having isolation trenches formed therein so as to remove defective regions, and isolation oxides formed in the isolation trenches.

Abstract: Speaker normalization is carried out based on biometric information available about a speaker, such as his height, or a dimension of a bodily member or article of clothing. The chosen biometric parameter correlates with the vocal tract length. Speech can be normalized based on the biometric parameter, which thus indirectly normalizes the speech based on the vocal tract length of the speaker. The inventive normalization can be used in model formation, or in actual speech recognition usage, or both. Substantial improvements in accuracy have been noted at little cost. The preferred biometric parameter is height, and the preferred form of scaling is linear scaling with the scale factor proportional to the height of the speaker.

Abstract: Disclosed is a method of automated handset identification, comprising receiving a sample speech input signal from a sample handset; deriving a cepstral covariance sample matrix from said first sample speech signal; calculating, with a distance metric, all distances between said sample matrix and one or more cepstral covariance handset matrices, wherein each said handset matrix is derived from a plurality of speech signals taken from different speakers through the same handset; and determining if the smallest of said distances is below a predetermined threshold value.

Abstract: There is provided a method for augmenting an alternate word list generated by a speech recognition system. The alternate word list includes at least one potentially correct word for replacing a wrongly decoded word. The method includes the step of identifying at least one acoustically confusable word with respect to the wrongly decoded word. The alternate word list is augmented with the at least one acoustically confusable word.

Abstract: A system for activating a microphone based on visual speech cues, in accordance with the invention, includes a feature tracker coupled to an image acquisition device. The feature tracker tracks features in an image of a user. A region of interest extractor is coupled to the feature tracker. The region of interest extractor extracts a region of interest from the image of the user. A visual speech activity detector is coupled to the region of interest extractor and measures changes in the region of interest to determine if a visual speech cue has been generated by the user. A microphone is turned on by the visual speech activity detector when a visual speech cue has been determined by the visual speech activity detector. Methods for activating a microphone based on visual speech cues are also included.

Abstract: An apparatus for automatically generating punctuation marks in a continuous speech recognition system, comprises means (1,2,3,5) for recognizing user speech and converting the user speech into words, characterized in that means (1,2,3,5) for recognizing user speech is further used to recognize pseudo noises in the user speech; and the apparatus characterized by further comprising: means (9) for marking pseudo noises in output results of means (1,2,3,5) for recognizing user speech; means (10,14,13) for generating punctuation marks by finding most likely pseudo punctuation marks at locations of pseudo noises marked by the means (9) for marking pseudo noises based on a language model containing pseudo punctuation marks.

Abstract: A system and method of controlling the operation of linear variable-gain amplifiers to allow for such linear variable gain amplifiers to have a wider operating range at high current levels, control inputs for selectable gains and improved low-voltage operation. In a first mode, the amplifier includes an additional source of current to allow for an enhanced operating range. In a second embodiment, the amplifier includes a plurality of selective resistive levels and a selection system which allows the selection of one of the resistive levels which, in turn, controls the gain range of the amplifier system of the present invention. A third embodiment of the present invention illustrates the use of an amplifier system useful for a low voltage input signal to reduce errors caused by variations in the base to emitter in the two transistors providing the amplification.