Abstract:
A data hiding method for hiding message data in media data and a data extracting method for extracting message data hidden in the media data. In the data hiding method, the message data are dispersedly hidden in the media data so that no third person can alter the message data. Specifically, the array elements of a message array are dispersedly hidden in a media array based on a status value S designating a specific array element in the media array when the media data and message data are expressed as the media array and the message array, respectively. The hiding method includes (a) a step of determining the j-th (j0) status value Sj, (b) a step of determining the (j+1)th status value Sj+1 based on the value Sj, the array element of the media array designated by the value Sj and the above-mentioned array element of the message array, and (c) a step of hiding data to be hidden in the array element of the media array designated by the value Sj+1.
Abstract:
PROBLEM TO BE SOLVED: To provide a computer-executable method for reducing dimensions, a program for making a computer execute the method, a dimension reduction device, and a search engine using the device. SOLUTION: The dimension reduction device which reduces the dimension of a numerical matrix using a computer to obtain a dimension-reduced matrix for providing information includes a processing part 32 which creates and stores a dimension-reduced matrix or index data for reducing dimensions, using a random average matrix RAV. The processing part 32 includes a shuffle vector creating part 44 for creating shuffle vectors for use as shuffle information, and a non-normal basic vector creating part 46 for creating a non-normal basic vector out of the numerical elements of data vectors designated by the shuffle vectors and for storing the non-normal basic vector. COPYRIGHT: (C)2005,JPO&NCIPI
Abstract:
PROBLEM TO BE SOLVED: To provide a method for retrieving information from a large scale database (containing several millions of data) in real time, while controlling trade-off in precision of a retrieved result and a response time thereof to a user. SOLUTION: This method is applicable, for example, to the database including a content of a document modeled in a clearly defined distance capable of calculating a distance between any two points, wherein the pair of documents near in its distance is more similar than a pair of documents apart each other. The method is combined with similarity ranking and/or another method to enhance scalability in information retrieval, detection, raking and tracking. COPYRIGHT: (C)2003,JPO
Abstract:
PROBLEM TO BE SOLVED: To provide an information retrieval system, an information retrieval method, a program for performing the information retrieval, and a recording medium for recording a program for performing the information retrieval. SOLUTION: This information retrieval system includes a means for producing and holding a specific matrix from a document-attribute matrix, a means for producing a document-attribute sub-matrix from a document added to the database for a predetermined period, a means for updating the specific matrix by using the information relating to the document-attribute sub-matrix, performing the singular value decomposition on the updated specific matrix, and reducing the dimension of all the document-attribute matrixes held in the database, and a means for performing the information retrieval by a query inputted by the user by using the document-attribute matrix of which the dimension reduced.
Abstract:
PURPOSE: To provide a stable speech synthesis processing with the reduced tremble of pitch in a speech synthesizer system utilizing a pitch synchronized waveform superposition method. CONSTITUTION: A glottis closing point is made to be the reference point of superposition (pitch mark). Since the glottis closing point is stably and accurately extracted by using a dynamic wavelet transformation, a speech with few trembling and few gurgling is synthesized by its stability. By setting the reference point of superposition on another position from the reference point of waveform cutout, the softer cutout of a wave form is enabled. Extraction of the glottis closing point is performed by searching the local peak of dynamic wavelet transformation, however, preferably the threshold for searching the local peak of dynamic wavelet transformation is adaptively controlled every time when dynamic wavelet transformation is obtained.
Abstract:
To provide a data hiding method of hiding media data in message data and a data extraction method of extracting the hidden data wherein message data is dispersively hidden in media data such as an image or sound to prevent a third person from modifying the message data easily. A data hiding method in which media data is expressed as a media array while message data is expressed as a message array so that the array elements of the message array can be dispersively hidden in the message array based on a state values specifying a particular array element of the media array, comprises the steps of: (a) determining the j-th (j ≥ 0) state value Sj; (b) determining (j + 1)-th state value Sj+1 based on the j-th state value, the array element of the media array indicated by the j-th state value, and the array element of the message array; and (c) hiding data with respect to the array element of the media array indicated by the (j + 1)-th state value Sj+1.