Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: A computer system is provided for training a neural network that converts images. Input images are applied to the neural network and a difference in image values is determined between predicted image data and target image data. A Fast Fourier Transform is taken of the difference. The neural network is trained on based the L1 Norm of resulting frequency data.

Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

Abstract: The provided solution makes it possible to select one item from a set of items, a multicomponent score being associated with each item of the set of items. After having computed a value characterising a relation ranking between a first item of the set of items and a second item of the set of items, the computed value being computed as a function of a plurality of score components associated with the first and second item of the set of items, respectively, the first item or of the second item is selected based on the computed value.

Abstract: The provided solution is an optical system comprising an image sensor and a collecting entrance lens having a predetermined field of view, the optical device further comprising a dynamically configurable light barrier having at least two masking elements, the dynamically configurable light barrier being configured so that only one of the masking elements can be set in a state in which light passes through, the at least two masking elements being configured so that images acquired by the image sensor when one of the at least two masking elements is configured in a state in which light passes through represent different portions of the predetermined field of view than images acquired by the image sensor when another one of the at least two masking elements is configured in a state in which light passes through.

Abstract: The present invention concerns the field of stereoscopic vision systems and more particularly the management of the position of the focal plane in a stereoscopic scene. The invention aims to propose a method for managing the focal plane within a stereoscopic image or a sequence of stereoscopic images. A translation of the right and left images is effected in order to place the focal plane in front of the scene viewed. Advantageously, the depth of the major objects of the scene is estimated by a division of the images into a plurality of zones and estimation of the depth of each of the zones. The depth of the focal plane is then calculated so that most of the scene is behind the focal plane.

Abstract: The present invention concerns the field of stereoscopic photography devices and more precisely calibration thereof. The invention proposes a method for calibrating a stereoscopic photography device that calculates a set of correction parameters for the images. These parameters are classed according to an order of importance. The first-order correction parameters are estimated first, the second-order correction parameters secondly. Advantageously, the first-order parameters are refined by taking account of the estimation values of the second-order parameters. Advantageously, a measurement of the relevance of the scene is carried out before the actual calibration.

Abstract: The provided solution makes it possible to select one item from a set of items, a multicomponent score being associated with each item of the set of items. After having computed a value characterising a relation ranking between a first item of the set of items and a second item of the set of items, the computed value being computed as a function of a plurality of score components associated with the first and second item of the set of items, respectively, the first item or of the second item is selected based on the computed value.

Abstract: The provided solution is an optical system comprising an image sensor and a collecting entrance lens having a predetermined field of view, the optical device further comprising a dynamically configurable light barrier having at least two masking elements, the dynamically configurable light barrier being configured so that only one of the masking elements can be set in a state in which light passes through, the at least two masking elements being configured so that images acquired by the image sensor when one of the at least two masking elements is configured in a state in which light passes through represent different portions of the predetermined field of view than images acquired by the image sensor when another one of the at least two masking elements is configured in a state in which light passes through.

Abstract: The present invention relates to the field of video compression and, specifically, to controlling the bit rate of the generated compressed sequence. This invention relates to a method for constant bit rate compression, using a step of normalizing the compression ratio. Variations in said ratio can thus be smoothed out, eliminating sudden variations in image quality in the resulting sequences. The perceived quality of the service is thereby improved.

Abstract: The present invention concerns the field of stereoscopic photography devices and more precisely calibration thereof. The invention proposes a method for calibrating a stereoscopic photography device that calculates a set of correction parameters for the images. These parameters are classed according to an order of importance. The first-order correction parameters are estimated first, the second-order correction parameters secondly. Advantageously, the first-order parameters are refined by taking account of the estimation values of the second-order parameters. Advantageously, a measurement of the relevance of the scene is carried out before the actual calibration.

Abstract: The present invention concerns the field of stereoscopic vision systems and more particularly the management of the position of the focal plane in a stereoscopic scene. The invention aims to propose a method for managing the focal plane within a stereoscopic image or a sequence of stereoscopic images. A translation of the right and left images is effected in order to place the focal plane in front of the scene viewed. Advantageously, the depth of the major objects of the scene is estimated by a division of the images into a plurality of zones and estimation of the depth of each of the zones. The depth of the focal plane is then calculated so that most of the scene is behind the focal plane.

Abstract: Audio encoding method and device comprising the transmission, in addition to the data representing a frequency-limited signal, of information relating to a temporal filter that can be applied to the entire broadened signal, both in its transmitted low-frequency part and in its reconstituted high-frequency part. The application of this filter allowing the reshaping the reconstituted high-frequency part and the correction of compression artifacts present in the transmitted low-frequency part. In this way, the application of the temporal filter, simple and inexpensive, to all or part of the reconstituted signal, makes it possible to obtain a signal of good perceived quality.

Abstract: Audio encoding method and device comprising the transmission, in addition to the data representing a frequency-limited signal, of information relating to a temporal filter that is to be applied to the entire enhanced signal, both in its transmitted low-frequency part and in its reconstituted high-frequency part. The application of this filter for reshaping the reconstituted high-frequency part and the correction of compression artefacts present in the transmitted low-frequency part. In this way, the application of the temporal filter, simple and inexpensive, to all or part of the reconstituted signal, makes it possible to obtain a signal of good perceived quality.

Abstract: The present invention relates to the field of video compression and, specifically, to controlling the bit rate of the generated compressed sequence. This invention relates to a method for constant bit rate compression, using a step of normalising the compression ratio. Variations in said ratio can thus be smoothed out, eliminating sudden variations in image quality in the resulting sequences. The perceived quality of the service is thereby improved.