Abstract: A method for speech recognition. The method uses a single pronunciation estimator to train acoustic phoneme models and recognize utterances from multiple languages. The method includes accepting text spellings of training words in a plurality of sets of training words, each set corresponding to a different one of a plurality of languages. The method also includes, for each of the sets of training words in the plurality, receiving pronunciations for the training words in the set, the pronunciations being characteristic of native speakers of the language of the set, the pronunciations also being in terms of subword units at least some of which are common to two or more of the languages. The method also includes training a single pronunciation estimator using data comprising the text spellings and the pronunciations of the training words.

Abstract: A method for speech recognition. The method uses a single pronunciation estimator to train acoustic phoneme models and recognize utterances from multiple languages. The method includes accepting text spellings of training words in a plurality of sets of training words, each set corresponding to a different one of a plurality of languages. The method also includes, for each of the sets of training words in the plurality, receiving pronunciations for the training words in the set, the pronunciations being characteristic of native speakers of the language of the set, the pronunciations also being in terms of subword units at least some of which are common to two or more of the languages. The method also includes training a single pronunciation estimator using data comprising the text spellings and the pronunciations of the training words.

Abstract: A method for speech recognition. The method uses a single pronunciation estimator to train acoustic phoneme models and recognize utterances from multiple languages. The method includes accepting text spellings of training words in a plurality of sets of training words, each set corresponding to a different one of a plurality of languages. The method also includes, for each of the sets of training words in the plurality, receiving pronunciations for the training words in the set, the pronunciations being characteristic of native speakers of the language of the set, the pronunciations also being in terms of subword units at least some of which are common to two or more of the languages. The method also includes training a single pronunciation estimator using data comprising the text spellings and the pronunciations of the training words.

Abstract: An input utterance is converted to a sequence of standard or prototype data frames which are compared with word models which are represented by respective sequences of standard or prototype probability states, there being a pre-calculable distance metric representing the degree of match between each prototype data frame and each prototype model state. Only distance measurements better than a calculated threshold are considered meaningful and those meaningful metrics are stored in a packed list. Also stored is an address array of offsets for locating particular meaningful metrics in the list, the address array being accessed by the corresponding frame and state indices. Also stored is an array for distinguishing meaningful and non-meaningful metrics. Accordingly, an input utterance can be evaluated by locating meaningful metrics in the packed list using the address array and by utilizing a default value for any non-meaningful metric.

Abstract: In the speech recognition system disclosed herein, the Viterbi decoding of an acoustic recognition network is augmented by implementing additional data structures for each arc in the network which represent, for each arc, the best path cost of reaching the last point on that arc and an arc in time which represents, for the best path, the time of leaving the previous arc. These additional data structures enable a trace back procedure which identifies not only the presumably optimal path but also alternate paths having good scores.

Abstract: An apparatus and allied method for detecting the presence of submerged, possibly distressed swimmers in a body of water employs a plurality of pairs of transducers arranged on opposite sides of the body of water. Pulsed sequential excitation of the transducers is employed to monitor the body of water. A person disposed between a pair of transducers interrupts the transmission of ultrasonic waves. An alarm is triggered upon the interruption of the ultrasonic waves or after a delay to avoid false alarms, warning of the presence and location of a submerged, lingering swimmer even in the presence of other active swimmers in the body of water. The same apparatus can be employed as an intrusion detector to detect unauthorized entry of a person into an unguarded body of water.

Abstract: A low conductivity fluid mask in a predetermined fluid state is used as a mask in an electroplating jet system method. Preferred embodiments describe a deionized water mask in the liquid fluid state.

Abstract: A parallel hybrid drive system for self propelled vehicles including a direct current motor-generator having a drive shaft in common with an internal combustion engine leading to a variable speed transmission and a final drive train. The motor-generator has a no-load speed, established by a separate field exciter, below which it drives the shaft as a motor and above which it is driven by the shaft as a generator. Storage batteries are operatively connected to the motor-generator to supply power to it below the no-load speed. The internal combustion engine operates over a small RPM range around its most efficient speed at a power level established by a fixed throttle setting. The transmission is operatively connected to an accelerator and direction selector switch to vary the speed and direction of the vehicle.

Abstract: A method for making lead frame devices wherein a thin metal foil is adhered to a stiffener panel by means of a photoresist layer to enable all processing steps by which the metal foil is transformed into an appropriate and predetermined lead frame configuration. After the lead frame fabrication is completed, the photoresist is removed by suitable means and ready for application and functional use as an electronic component electrical connecting device.