Abstract: An information processing method includes: reading a layer structure and parameters of layers from each of models of two neural networks; and determining a degree of matching between the models of the two neural networks, by comparing layers, of the respective models of the two neural networks, that are configured as a graph-like form in respective hidden layers, in order from an input layer using breadth first search or depth first search, based on similarities between respective layers.

Abstract: An information processing method for embedding watermark bits into weights of a first neural network includes: obtaining an output of a second neural network by inputting a plurality of input values obtained from a plurality of weights of the first neural network to the second neural network; obtaining second gradients of the respective plurality of input values based on an error between the output of the second neural network and the watermark bits; and updating the weights based on values obtained by adding first gradients of the weights of the first neural network that have been obtained based on backpropagation and the respective second gradients.

Abstract: An information processing method for embedding watermark bits into weights of a first neural network includes: obtaining an output of a second neural network by inputting a plurality of input values obtained from a plurality of weights of the first neural network to the second neural network; obtaining second gradients of the respective plurality of input values based on an error between the output of the second neural network and the watermark bits; and updating the weights based on values obtained by adding first gradients of the weights of the first neural network that have been obtained based on backpropagation and the respective second gradients.

Abstract: An information processing method includes: reading a layer structure and parameters of layers from each of models of two neural networks; and determining a degree of matching between the models of the two neural networks, by comparing layers, of the respective models of the two neural networks, that are configured as a graph-like form in respective hidden layers, in order from an input layer using breadth first search or depth first search, based on similarities between respective layers.

Abstract: A method of providing a video content formed of a plurality of successive video frames is disclosed. The method includes: extracting a visual feature from an image of each of subject frames which are all or some of the plurality of video frames; and selecting one of a plurality of candidate overlay insertion regions which are pre-assigned on each of the subject frames at different positions for allowing an additional content as a content overlay to be overlaid on the subject frames for display, as a final overlay insertion region, based on distances of the plurality of candidate overlay insertion regions from the extracted feature.

Abstract: The present invention relate to an image generating apparatus for generating an image from a viewpoint specified by a user. According to the invention, the apparatus has a storage unit that stores data of a plurality of images, and a disparity map generating unit that generates a disparity map. The disparity map is associated with a pair of images including a first image and a second image, and indicates corresponding pixels in the first and second images.

Abstract: It is a purpose of the present invention to provide a video encoding apparatus and a video decoding apparatus which are capable of providing real-time processing having improved coding performance. A video encoding apparatus AA includes an inter-encoding predicted value generating unit 2 configured to receive, as input signals, an input image a, a local decoded value d acquired from an encoded block, prediction direction information e acquired from the encoded block, and interpolation filter initial control information b. The inter-encoding predicted value generating unit 2 includes a filter control unit 21. The filter control unit 21 extracts the tap size N of the interpolation filter and the initial value c0(i,j) of the filter coefficient of the interpolation filter from the interpolation filter initial control information b, and uses the initial value c0(i,j) of the filter coefficient of the interpolation filter as the filter coefficient c(i,j) of the interpolation filter.

Abstract: An objective image quality assessment device of a video quality for estimating a subjective image quality from compressed bit stream is provided. The device includes a macro-block layer parameter analyzer for acquiring a quantizer scale size defined with respect to each macro-block and a coding type of a slice from a compressed bit stream, an orthogonal transformation coefficient in each macro-block, and a reference frame number in the macro-block applied with a motion compensating prediction, an average quantizer scale calculator for obtaining an average in a sequence of the quantizer scale size for every coding type of the slice, a spatial degradation feature calculator and a temporal degradation feature calculator for obtaining a spatial degradation feature and a temporal degradation feature, in each macro-block, and an image feature integrator for deriving the objective image quality based on the features obtained in the calculators.

Abstract: It is a purpose of the present invention to provide a video encoding apparatus, a video decoding apparatus, a video encoding method, a video decoding method, and a computer program, which allow adaptive filter coefficients to be used with a reduced amount of coding. A video encoding apparatus AA selects a suitable coefficient from among multiple kinds of interpolation filter coefficients determined beforehand by means of a component external to the video encoding apparatus AA. Subsequently, the video encoding apparatus AA records identification information uniquely defined for the selected interpolation filter coefficient in control information for each slice. Furthermore, the video encoding apparatus AA uses the interpolation filter coefficient thus selected to generate encoded data c including the identification information uniquely defined for the interpolation filter coefficient thus selected.

Abstract: A barcode-embedding area is obtained in a color image based on expression possibility by the juxtaposed additive color mixture and possibility of reading of pixels separately by a barcode reader (S11), a color image in the barcode-embedding area is expressed by the juxtaposed additive color mixture, and color images for the code area and the background area, which are readable by the barcode reader, are created (S12, S13), a color-image-superimposed barcode image is created by using the color images for the code area and the background area, and the same color-image-superimposed barcode image is embedded in the barcode-embedding area (S14), and the color-image-superimposed barcode image and the original color image are corrected if necessary (S15, S16).

Abstract: A feature amount extracting unit extracts a block distortion degree of an evaluated video y relative to an original video x, a PSNR overall temporal fluctuation degree, and a PSNR local temporal fluctuation degree as feature amounts. A weighted sum calculating unit calculates a weighted sum of these feature amounts, and calculates an objective video quality index. Frame rate detecting unit detects frame rate of the evaluated video y. A correcting unit corrects the objective video quality index based on the frame rate detected by the frame rate detecting unit. An objective evaluation index-subjective video quality mapping unit applies the corrected objective video quality index Qobj to a correlation between an objective video quality index and a subjective video quality given in advance, thereby deriving a subjective video quality estimated value DMOS.

Abstract: The present invention relates to color correction of video signals from a plurality of cameras. The apparatus has a correction order that defines a plurality of camera pairs and an order of the camera pairs, where each camera pair specifies a reference camera and a target camera. In the correction order, each camera is treated as a target camera at least one time, and the reference camera is treated as the target camera in a previous camera pair except for the first camera pair. The apparatus corrects a value of each pixel in a target image captured by the target camera based on a reference image captured by the reference camera, and replaces the target image by the corrected target image, for each camera pair in accordance with the order of the camera pairs.

Abstract: The present invention provides a method for generating free viewpoint video image in three-dimensional movement capable of synthesizing the free viewpoint video image from a viewpoint which looks down on an object from above. This method includes a process of taking multi-viewpoint video images using a plurality of cameras located on an identical plane and a camera not located on the identical plane, a process of generating video image at a viewpoint having the same azimuth as a desired viewpoint and located on the plane from the multi-viewpoint video images of the cameras on the plane, and a process of generating video image at the desired viewpoint from video image of the camera not located on the plane and video image at the viewpoint having the same azimuth as the desired viewpoint and located on the plane.

Abstract: The present invention provides a method of generating a virtual viewpoint video image when the virtual viewpoint position is not located on a plane where a camera is disposed. In an environment in which a plurality of cameras having a horizontal optical axis are disposed in a real zone (for example, on the circumference) which surrounds an object, a video image of an arbitrary viewpoint on the circumference is generated. Further, by synthesizing video images photographed by a camera, a free viewpoint video image is generated from a virtual viewpoint (viewpoint from a high or low position) where no camera is placed. According to a method of achieving this, a travel distance of a display position is calculated by the local region synthesizing portion and this travel distance is reflected to the free viewpoint video image of a local region.

Abstract: The present invention relates to an apparatus and a method for classifying pixels in each frame of a motion picture as foreground or background. According to the invention, the apparatus has a decision unit and adjustment unit. The decision unit classifies a pixel as a foreground pixel, a background pixel or an undefined pixel based on a first threshold, a second threshold and a distribution value of the pixel representing an occurrence probability of a pixel value of the pixel. The decision unit sorts each pixel into a foreground pixel or a background pixel based on the classification result of the decision unit.

Abstract: The present invention provides a scrambled compressed data distribution device that can perform a high-resistance scrambled compressed data distribution by selectively using a plurality of types of scramble information and a playback device thereof. A scrambling section scrambles compressed data by unique one of a plurality of scrambling methods. A hash value calculating section determines a hash value based on data b in a position where digital watermark data is not embedded, and an encrypting section encrypts the hash value with a secret key Sk(n) corresponding to the scrambling method and outputs an encrypted hash value c. A watermark data embedding section embeds the encrypted hash value in a watermark data embedding position of the data b and transmits the same as a compressed data stream d.

Abstract: The present invention relates to calibration of camera parameters for converting a world coordinate system, which indicates a position in the real space, to a coordinate used in an image and vice versa. The apparatus according to the invention has a detection unit, which determines corresponding pixel pairs from the captured image and the model image and outputs corresponding data indicating determined pixel pairs, and a selection unit, which selects pixel pairs to be left in the corresponding data and removes data related to an unselected pixel pair from the corresponding data for generating selected corresponding data. The apparatus further has a calculation unit, which calculates camera parameters based on the selected corresponding data.

Abstract: It is a purpose of the present invention to provide a video encoding apparatus, a video decoding apparatus, a video encoding method, a video decoding method, and a computer program, which allow adaptive filter coefficients to be used with a reduced amount of coding. A video encoding apparatus AA selects a suitable coefficient from among multiple kinds of interpolation filter coefficients determined beforehand by means of a component external to the video encoding apparatus AA. Subsequently, the video encoding apparatus AA records identification information uniquely defined for the selected interpolation filter coefficient in control information for each slice. Furthermore, the video encoding apparatus AA uses the interpolation filter coefficient thus selected to generate encoded data c including the identification information uniquely defined for the interpolation filter coefficient thus selected.

Abstract: A method of providing a video content formed of a plurality of successive video frames is disclosed. The method includes: extracting a visual feature from an image of each of subject frames which are all or some of the plurality of video frames; and selecting one of a plurality of candidate overlay insertion regions which are pre-assigned on each of the subject frames at different positions for allowing an additional content as a content overlay to be overlaid on the subject frames for display, as a final overlay insertion region, based on distances of the plurality of candidate overlay insertion regions from the extracted feature.

Abstract: The present invention provides a scrambled compressed data distribution device that can perform a high-resistance scrambled compressed data distribution by selectively using a plurality of types of scramble information and a playback device thereof. A scrambling section scrambles compressed data by unique one of a plurality of scrambling methods. A hash value calculating section determines a hash value based on data b in a position where digital watermark data is not embedded, and an encrypting section encrypts the hash value with a secret key Sk(n) corresponding to the scrambling method and outputs an encrypted hash value c. A watermark data embedding section embeds the encrypted hash value in a watermark data embedding position of the data b and transmits the same as a compressed data stream d.