Abstract: A conversational agent capability assessment method, system, and computer program product, includes obtaining data to create at least one scenario for testing a conversational agent, performing a set of tests using a scenario of the at least one scenario created to assess a capability of the conversational agent, and comparing a result of the capability from the set of tests with an expected result of the scenario.

Abstract: A deburring device includes a deburring tool for removing burrs from an object, a robot for moving an object or the tool, a force sensor for detecting force acting on the tool, and a visual sensor for detecting a position of a burr portion of the object. According to the deburring device, information regarding shape data of the burr portion and a posture of the tool is obtained beforehand based on three-dimensional data of the object. Based on the shape data and the posture of the tool, a robot program is created. In accordance with an actual burr portion detected by the visual sensor, the robot program is replaced as necessary. During the deburring, the robot is controlled according to the force control by using a detected value from the force sensor.

Abstract: Techniques include performing text mining on a set of case reports in text format to determine a set of grammar rules to be used to determine whether case reports meet a medical condition. The text mining includes performing feature selection, used to determine the set of grammar rules, that combines standardized case definitions with experience of medical officers for the medical condition and outputting the set of grammar rules. Another technique includes applying grammar rule(s) to new case report(s), the grammar rule(s) previously determined at least by performing text mining comprised of performing feature selection, used to determine the set of grammar rules, that combines standardized case definitions with experience of medical officers for the medical condition. Indication(s) are output of whether the new case report(s) meet or do not meet the medical condition. The techniques may be performed by a method, an apparatus, and a program product.

Abstract: A method, system, and computer program product for creating confidence-rated transcription and translation are provided in the illustrative embodiments. An input is provided in a first form to a set of transcription applications. A set of transcriptions is received. A first and a second set of confidence ratings are assigned to a first and a second transcription, respectively. The confidence-rated first transcription and the confidence-rated second transcription are combined and provided to a set of translation applications. A set of translations is received. A third and a fourth set of confidence ratings are assigned to a first and a second translation, respectively. The confidence-rated first and second translations are combined and presented.

Abstract: In automated speech recognition (ASR), multiple devices may be employed to perform the ASR in a distributed environment. To reduce bandwidth use in transmitting between devices ASR information is compressed prior to transmission. To counteract fidelity loss that may accompany such compression, two versions of an audio signal are processed by an acoustic front end (AFE), one version is unaltered and one is compressed and decompressed prior to AFE processing. The two versions are compared, and the comparison data is sent to a recipient for further ASR processing. The recipient uses the comparison data and a received version of the compressed audio signal to recreate the post-AFE processing results from the received audio signal. The result is improved ASR results and decreased bandwidth usage between distributed ASR devices.

Abstract: A high quality speech is reproduced with a small data amount in speech coding and decoding for performing compression coding and decoding of a speech signal to a digital signal. Ina speech coding method according to a code-excited linear prediction (CELP) speech coding, a noise level of a speech in a concerning coding period is evaluated by using a code or coding result of at least one of spectrum information, power information, and pitch information, and various excitation code books are used based on an evaluation result.

Abstract: A speech processing system which exploits statistical modeling and formal logic to receive and process speech input, which may represent data to be received, such as dictation, or commands to be processed by an operating system, application or process. A command dictionary and dynamic grammars are used in processing speech input to identify, disambiguate and extract commands. The logical processing scheme ensures that putative commands are complete and unambiguous before processing. Context sensitivity may be employed to differentiate data and commands. A multi faceted graphic user interface may be provided for interaction with a user to speech enable interaction with applications and processes that do not necessarily have native support for speech input.

Abstract: A speech synthesis device is provided with: a central segment selection unit for selecting a central segment from among a plurality of speech segments; a prosody generation unit for generating prosody information based on the central segment; a non-central segment selection unit for selecting a non-central segment, which is a segment outside of a central segment section, based on the central segment and the prosody information; and a waveform generation unit for generating a synthesized speech waveform based on the prosody information, the central segment, and the non-central segment. The speech synthesis device first selects a central segment that forms a basis for prosody generation and generates prosody information based on the central segment so that it is possible to sufficiently reduce both concatenation distortion and sound quality degradation accompanying prosody control in the section of the central segment.

Abstract: An encoding apparatus and method for generating low-frequency-band encoding information and high-frequency-band encoding information from an original signal. The encoding apparatus includes a first spectrum calculator that calculates a first spectrum of a low frequency band from a decoded signal of the low-frequency-band encoding information, a second spectrum calculator that calculates a second spectrum from the original signal, an estimator that divides a high frequency band of the second spectrum into a plurality of bands and estimates the second spectrum included in each band, using the first spectrum, and a first error component encoder that encodes a first error component between the high frequency band of the second spectrum and an estimated spectrum. A corresponding decoding apparatus and method provides decoding.