The present invention generally relates to a method for predicting a block of pixels from at least one patch comprising a block of pixels and a causal neighborhood around this block of pixels. The method comprises the following steps: determining a mapping of a causal neighborhood, around that block of pixels to be predicted, on the block of pixels to be predicted in order that the block of pixels of each patch is best predicted by mapping the neighborhood of that patch on the block of pixels of that patch, and predicting the block of pixels from a prediction block computed by applying the determined mapping on the neighborhood of the block of pixels to predict.

Methods and apparatus are provided for sparsity-based de-artifact filtering for video encoding and decoding. An apparatus includes an encoder (400) for encoding at least a portion of an image by grouping regions within the portion based on a grouping metric, transforming the grouped regions, adaptively performing de-artifact filtering on the transformed regions using a de-artifacting filter (413) included in the encoder, inverse transforming the de-artifacted regions to create replacement regions, and restoring the replacement regions to positions with the image from which the regions were taken prior to the grouping.

When the signal format of a color image is YUV4:2:2, and an intra prediction parameter, which is used for an intra-frame prediction process on the chrominance signals in a prediction block, shows the same prediction mode as an intra prediction mode for the luminance signal in the prediction block, an intra prediction unit 4 converts an index indicating the intra prediction mode for the luminance signal, and performs the intra-frame prediction process on the chrominance signals in an intra prediction mode indicated by the index after conversion.

Among other things, one or more techniques and/or systems are provided for defining a view direction for a texture image used to texture a geometry. That is, a geometry may represent a multi-dimensional surface of a scene, such as a city. The geometry may be textured using one or more texture images depicting the scene from various view directions. Because more than one texture image may contribute to texturing portions of the geometry, a view direction for a texture image may be selectively defined based upon a coverage metric associated with an amount of non-textured geometry pixels that are textured by the texture image along the view direction. In an example, a texture image may be defined according to a customized configuration, such as a spherical configuration, a cylindrical configuration, etc. In this way, redundant texturing of the geometry may be mitigated based upon the selectively identified view direction(s).

A method for encoding at least one extra bit in an image compression and decompression system. The method includes accessing an input image, and compressing the input image into a compressed image using an encoder system, wherein said encoding system implements an algorithm for encoding at least one extra bit. The method further includes communicatively transferring the compressed image to a decoding system, and decompressing the compressed image into a resulting uncompressed image that is unaltered from said input image, wherein the algorithm for encoding enables the recovery of the at least one extra bit.

Systems and methods are provided for encoding raw image content encoded using one or more virtual intra-frames. In an exemplary method, a stream of video content may be encoded to generate compressed video data that includes an intra-frame and a plurality of corresponding inter-frames. The compressed video data may be stored within a buffer, and when the amount of data in the buffer exceeds a threshold value, a virtual intra-frame may be computed based on decoded inter-frame data. The virtual intra-frame may be output for storage in the buffer.

The invention relates to a method for selecting an image dynamic range conversion operator from among a set of image dynamic range conversion operators by optimizing a quality criterion, each image dynamic range conversion operator being applied to an original image to obtain an image, called the modified image, whose dynamic range is lower than that of the original image, wherein said quality criterion is calculated, for each image dynamic range conversion operator, as a function of a distortion calculated from a reconstructed version after coding and decoding the original image modified by said image dynamic range conversion operator.

New intra planar modes are introduced for predicting digital video data. As part of the new intra planar modes, various methods are offered for predicting a first sample within a prediction unit, where the first sample is needed for referencing to when processing the new intra planar modes. And once the first sample is successfully predicted, the new intra planar modes are able to predict a sample of video data within the prediction unit by processing a bi-linear interpolation of four previously reconstructed reference samples.

Disclosed herein is a method of processing a graph-based signal using a geometric primitive, comprising: specifying the geometric primitive to be used for calculating an edge weight; obtaining a parameter for each of the geometric primitive; calculating an edge weight for each of edges within the image based on the parameter; and encoding the image based on the edge weight.

A method for training a face detection model, including: acquiring a training face image; performing three-dimensional reconstruction on the training face image based on a preset three-dimensional face model, and acquiring a training three-dimensional face model; generating a training UV coordinate map including three-dimensional coordinates of the training three-dimensional face model based on the training three-dimensional face model; and training a semantic segmentation network by using the training face image and the training UV coordinate map, and acquiring a face detection model. The face detection model is configured to generate a UV coordinate map including three-dimensional coordinates.