Abstract: A speech recognition apparatus includes a word dictionary having recognition target words, a first acoustic model which expresses a reference pattern of a speech unit by one or more states, a second acoustic model which is lower in precision than said first acoustic model, selection means for selecting one of said first acoustic model and said second acoustic model on the basis of a parameter associated with a state of interest, and likelihood calculation means for calculating a likelihood of an acoustic feature parameter with respect to said acoustic model selected by said selection means.

Abstract: Input text data undergoes language analysis to generate prosody, and a speech database is searched for a synthesis unit on the basis of the prosody. A modification distortion of the found synthesis unit, and concatenation distortions upon connecting that synthesis unit to those in the preceding phoneme are computed, and a distortion determination unit weights the modification and concatenation distortions to determine the total distortion. An Nbest determination unit obtains N best paths that can minimize the distortion using the A* search algorithm, and a registration unit determination unit selects a synthesis unit to be registered in a synthesis unit inventory on the basis of the N best paths in the order of frequencies of occurrence, and registers it in the synthesis unit inventory.

Abstract: In a system implementing image retrieval by performing speech recognition on voice information added to an image, the speech recognition is triggered by an event, such as an image upload event, that is not an explicit speech-recognition order event. The system obtains voice information added to an image, detects an event, and performs speech recognition on the obtained voice information in response to a specific event, even if the detected event is not an explicit speech-recognition order event.

Abstract: An amplitude altering magnification (r) applied to sub-phoneme units of a voiced portion and an amplitude altering magnification s to be applied to sub-phoneme units of an unvoiced portion are determined based upon a target phoneme average power (p0) of synthesized speech and power (p) of a selected phoneme unit. Sub-phoneme units are extracted from a phoneme to be synthesized. From among the extracted sub-phoneme units, a sub-phoneme unit of the voiced portion is multiplied by the amplitude altering magnification (r), and a sub-phoneme unit of the unvoiced portion is multiplied by the amplitude altering magnification (s). Synthesized speech is obtained using the sub-phoneme units thus obtained. This makes it possible to realize power control in which any decline in the quality of synthesized speech is reduced.

Abstract: A signal processing apparatus and method for performing a robust endpoint detection of a signal are provided. An input signal sequence is divided into frames each of which has a predetermined time length. The presence of the signal in the frame is detected. After that, the filter process of smoothing the detection result by using the detection result for a past frame is applied to the detection result for a current frame. The filter output is compared with a predetermined threshold value to determine the state of the signal sequence of the current frame on the basis of the comparison result.

Abstract: A speech segment search unit searches a speech database for speech segments that satisfy a phonetic environment, and a HMM learning unit computes the HMMs of phonemes on the basis of the search result. A segment recognition unit performs segment recognition of speech segments on the basis of the computed HMMs of the phonemes, and when the phoneme of the segment recognition result is equal to a phoneme of the source speech segment, that speech segment is registered in a segment dictionary.

Abstract: In cases where at least one item of sound information has been associated with at least image, at least one desired item of sound information is selected and the sound information is played back in a prescribed order. According, in an information processing apparatus, a playback sequence decision unit (103) reads in image data as well as sound data, which has been assigned within the image data, from a image/sound data storage unit (107), generates a still image in which the positions at which sound data has been recorded is denoted on the image, and displays the generated still image on a image display unit (106). A sound data specifying unit (102) searches the image/sound data storage unit (107) for sound data that has been associated with the interior of an image area specified by an input from a user. When applicable sound data is found to exist, the playback sequence decision unit (103) decides the order in which the applicable sound data is to be played back.

Abstract: An article storage apparatus which enables categories to be subsequently assigned to storage sections or categories given to storage sections to be changed according to the progress of the user's storing operation or the user's desire. A plurality of storage shelves are provided to store articles. A RFID reader reads out a category assigned to an article to be stored in each of the storage shelves. A controller sets a category to be assigned to each storage shelf according to the category assigned to the article stored in the storage shelf.

Abstract: More comfortable data input is implemented by using speech recognition and a character prediction function in combination. For example, according to a data input method of this invention, character string candidates which follow a character string input by a character string input device are predicted (S402), and the predicted character string candidates are displayed on a display device (S403). Speech recognition is performed for speech input by the speech input device using the character string candidates displayed on the display device as words to be recognized (S411), and a character string serving as the recognition result is confirmed as a character string to be used (S412).

Abstract: Robust signal detection against various types of background noise is implemented. According to a signal detection apparatus and method of this invention, the feature amount of an input signal sequence and the feature amount of a noise component contained in the signal sequence are extracted. After that, the first likelihood indicating probability that the signal sequence is detected and the second likelihood indicating probability that the noise component is detected are calculated on the basis of a predetermined signal-to-noise ratio and the extracted feature amount of the signal sequence. Additionally, a likelihood ratio indicating the ratio between the first likelihood and the second likelihood is calculated. Detection of the signal sequence is determined on the basis of the likelihood ratio.

Abstract: An information processing technique for voice outputting an electronic mail, received by an information processing apparatus capable of voice output, at a sender's intended timing. For this purpose, the information processing apparatus has an electronic mail reception unit (101) to receive an electronic mail, an electronic mail selection unit (102) to select an electronic mail including a code describing voice output timing, from electronic mails received by the electronic mail reception unit (101), and a voice synthesis unit (104) to voice-synthesize the electronic mail selected by the electronic mail selection unit (102) and voice-outputs the result of voice synthesis based on the code.

Abstract: The present invention is purposed to improve user friendliness in generation of practical data access means. To achieve this object, the present invention provides a data processing method of registering a path for data access and link data for the path. The method comprises: a generation step of the link data candidate generation unit 202 for generating a link data candidate based on a file accessed from the path which is inputted for data access; a display step of the link data candidate exhibiting unit 203 for displaying the generated link data candidate; a recognition step of the link data selection unit 204 for recognizing link data selected from the displayed link data candidate; and a registration step of the link data registration unit 205 for registering the recognized link data in association with the path of the accessed file.

Abstract: The system implements high-accuracy speech recognition while suppressing the amount of data transfer between the client and server. For this purpose, the client compression-encodes speech parameters by a speech processing unit, and sends the compression-encoded speech parameters to the server. The server receives the compression-encoded speech parameters, a speech processing unit makes speech recognition of the compression-encoded speech parameters, and sends information corresponding to the speech recognition result to the client.

Abstract: Input text data undergoes language analysis to generate prosody, and a speech database is searched for a synthesis unit on the basis of the prosody. A modification distortion of the found synthesis unit, and concatenation distortions upon connecting that synthesis unit to those in the preceding phoneme are computed, and a distortion determination unit weights the modification and concatenation distortions to determine the total distortion. An Nbest determination unit obtains N best paths that can minimize the distortion using the A* search algorithm, and a registration unit determination unit selects a synthesis unit to be registered in a synthesis unit inventory on the basis of the N best paths in the order of frequencies of occurrence, and registers it in the synthesis unit inventory.

Abstract: A sheet which gives an absorbent article improved recovery from distortion is disclosed. The sheet has a recovery force of 0.7 cN or more in the cross direction, a compressive strength of 100 cN or less, and a basis weight of 20 g/m2 or more.

Abstract: The system implements high-accuracy speech recognition while suppressing the amount of data transfer between the client and server. For this purpose, the client compression-encodes speech parameters by a speech processing unit, and sends the compression-encoded speech parameters to the server. The server receives the compression-encoded speech parameters, and speech processing unit makes speech recognition of the compression-encoded speech parameters, and sends information corresponding to the speech recognition result to the client.

Abstract: A method for restoring bulkiness of nonwoven fabric which contains crimped thermoplastic fiber and is in a roll form is disclosed. The method comprises unwinding the nonwoven fabric from the stock roll, and blowing hot air to the unwound nonwoven fabric by a through-air technique to make the nonwoven fabric increase in bulkiness. The hot air is heated at a temperature lower than the melting point of the thermoplastic fiber and not lower than a temperature lower than that melting point by about 50° C., and is blown for about 0.05 to 3 seconds.

Abstract: An information processing device according to the present invention is included in an information processing apparatus and outputs guidance information for an operation performed for the information processing apparatus by a user. In the information processing device, a user information acquisition unit identifies a user who is operating the information processing device, and an input control unit identifies the type of operation performed by the user. The information processing device also includes an operation history database for storing operation history information unique to the user and a voice guidance database for storing at least one piece of guidance information on the operation. A guidance selection unit selects appropriate guidance information on the basis of the operation history information on the operation unique to the user, and a voice output unit outputs the selected guidance information.

Abstract: In a speech synthesis process, micro-segments are cut from acquired waveform data and a window function. The obtained micro-segments are re-arranged to implement a desired prosody, and superposed data is generated by superposing the re-arranged micro-segments, so as to obtain synthetic speech waveform data. A spectrum correction filter is formed based on the acquired waveform data. At least one of the waveform data, micro-segments, and superposed data is corrected using the spectrum correction filter. In this way, “blur” of a speech spectrum due to the window function applied to obtain micro-segments is reduced, and speech synthesis with high sound quality is realized.

Abstract: Detecting an unknown word in input speech data reduces the search space and the memory capacity for the unknown word. For this purpose, an HMM data memory stores data describing a state transition mode for the unknown word, defined by a number of states and the transition probability between the states. An output probability calculation unit acquires a state of the maximum likelihood at each time of the speech data, among the plural states employed in the state transition mode for a known word, employed in the speech recognition of the known word. The obtained result is applied to the state transition mode for the unknown word, stored in the HMM data memory, to obtain a state transition mode of the unknown word. A different output probability calculation unit determines the likelihood of the state transition mode for the known word.