Abstract: Methods and apparatus of video coding using sub-block based affine mode are disclosed. According to this method, control-point motion vectors (MVs) associated with the affine mode are determined for a block. A sub-block MV is derived for a target sub-block of the block from the control-point MVs for the block. A prediction offset is determined for a target pixel of the target sub-block using information comprising a pixel MV offset from the sub-block MV for the target pixel according to Prediction Refinement with Optical Flow (PROF). The target pixel of the target sub-block is encoded or decoded using a modified predictor. The modified prediction is generated by clipping the prediction offset to a target range and combining the clipped prediction offset with an original predictor.

Abstract: At least a method and an apparatus are presented for efficiently encoding or decoding video. For example, a plurality of frames is provided to a motion estimator to produce an output comprising estimated motion information. The estimated motion information is provided to an auto-encoder or an auto-decoder to produce an output comprising reconstructed motion field. The reconstructed motion field and one or more decoded frames of the plurality of frames are provided to a deep neural network to produce an output comprising refined bi-directional motion field. The video is encoded or decoded based on the refined bi-directional motion field.

Abstract: Methods, systems, apparatus for video processing are described. A method for processing video includes performing a conversion between a video and a bitstream of the video according to a format rule, and wherein the format rule specifies whether a syntax element in a sequence parameter set (SPS) based on a value of a general constraint flag.

Abstract: An image encoding/decoding method and apparatus are provided. The image decoding method includes determining whether a prediction mode of a current block is an intra prediction mode based on information on the prediction mode of the current block, determining whether a transform mode of the current block is a transform skip mode, determining whether an intra prediction technique of the current block is intra sub-partitions (ISP) based on the prediction mode of the current block being an intra prediction mode, and generating a transform coefficient block of the current block based on the transform mode of the current block and whether the intra prediction technique of the current block is ISP.

Abstract: An aerial vehicle includes a vehicle body coupled to one or more blades configured to provide lift to the aerial vehicle, an image device including a first housing containing an optical unit and a second housing containing a non-optical unit and not the optical unit, an apparatus connected to the vehicle body and configured to stabilize at least a portion of the imaging device. The optical unit includes a lens and a photosensor. The apparatus includes a frame assembly, an inertial measurement unit, and a motor assembly coupled to the frame assembly. The frame assembly is configured to support the optical unit without supporting the non-optical unit, and to permit rotation of the optical unit about a first rotational axis or a second rotational axis. The frame assembly includes a first frame member supporting the optical unit, and a second frame member bearing weight of the first frame member.

Abstract: A set of tensor-product B-Spline (TPB) basis functions is determined. A set of selected TPB prediction parameters to be used with the set of TPB basis functions for generating predicted image data in mapped images from source image data in source images of a source color grade is generated. The set of selected TPB prediction parameters is generated by minimizing differences between the predicted image data in the mapped images and reference image data in reference images of a reference color grade. The reference images correspond to the source images and depict same visual content as depicted by the source images. The set of selected TPB prediction parameters is encoded in a video signal as a part of image metadata along with the source image data in the source images. The mapped images are caused to be reconstructed and rendered with a recipient device of the video signal.

Abstract: A method of decoding a coded video bitstream implemented by a video decoder is provided. The method includes receiving, by a receiver of the video decoder, a mixed intra random access point (IRAP) picture including a first sub-picture and a second sub-picture, wherein the first sub-picture is an IRAP picture and the second sub-picture is a non-IRAP sub-picture; receiving, by the receiver; a reference picture list (RPL) for the mixed IRAP picture; decoding, by a processor of the video decoder, the second sub-picture using the RPL; and generating, by the processor, an image based on the second sub-picture as decoded. A corresponding encoding method is also provided.

Abstract: A method of video processing is provided to include: deciding a coding mode used for representing a current video block of a video in a coded representation of the video; and coding the current video block into the coded representation according to the coding mode, wherein use of alternative half-pel accuracy filters in addition to a default half-pel accuracy filter for representing motion information is disabled for the current video block due to use of the coding mode.

Abstract: Methods, systems, apparatus for video processing are described. A method for processing video includes performing a conversion between a video and a bitstream of the video according to a format rule, and wherein the format rule specifies whether a syntax element in a sequence parameter set (SPS) based on a value of a general constraint flag.

Abstract: Restrictions of usage of non-power-two-partition tree in video compression is described. In an exemplary aspect, a method for video processing includes determining, for a conversion between a video and a bitstream representation of the video, whether non-power-of-two partition trees (NPT-T) partition is enabled or disabled, wherein the NPT-T partition include splitting a first block of the video into multiple smaller sized child blocks of the first block, and width (Wi) and/or height (Hi) of at least one child block is a non-power-of-two integer; determining restrictions associated with usage of the NPT-T partition in response to the determination that the NPT-T partition is allowed; and performing the conversion based on the determinations.

Abstract: In encoding a sequence of frames, a linear transform, such as, for example a wavelet transform, is applied to the sequences of frames on a frame-by-frame basis. At least one portion of a linearly transformed frame thus obtained is encoded in the inter mode on the basis of a differential representation of the linearly transformed frame. The differential representation corresponds with a difference between, on the one hand, the portion of the linearly transformed frame and, on the other hand, a representation of a corresponding portion of at least one other linearly transformed frame. Data compression is then applied to the differential representation of the portion of the linearly transformed frame.

Abstract: A video encoding method of a cloud game is provided. Network bandwidth information of a video playback terminal is determined. Decoding computing power information of the video playback terminal is determined. At least one encoding parameter of a video of the cloud game is dynamically adjusted based on (i) the network bandwidth information and (ii) the decoding computing power information. The network bandwidth information and the decoding computing power information indicate a playback environment of the video of the cloud game. The video of the cloud game is encoded based on the at least one encoding parameter.

Abstract: A video encoder is disclosed. The video encoder comprises an integer level motion estimation hardware component configured to determine candidate integer level motion vectors for a video being encoded. The video encoder comprises a fractional motion estimation hardware component configured to receive the candidate integer level motion vectors from the integer level motion estimation hardware component and refine the candidate integer level motion vectors into candidate sub-pixel level motion vectors. The fractional motion estimation hardware component includes parallel pipelines configured to process coding units of a frame of the video in parallel across the parallel pipelines.

Abstract: The technology described herein relates to online per-title encoding. A method for online per-title encoding includes receiving a video input, generating segments of the video input, extracting a spatial feature and a temporal feature, predicting bitrate-resolution pairs based on the spatial feature and the temporal feature, using a discrete cosine transform (DCT)-based energy function, and per-title encoding segments of the video input for the predicted bitrate-resolution pairs. A system for online per-title encoding may include memory for storing a set of bitrates, a set of resolutions, and a machine learning module configured to predict bitrate resolution pairs based on low-complexity spatial and temporal features.

Abstract: A method of video processing includes determining a target bitrate for a current video unit of a video that is based on a rate distortion function in which a rate portion is weighted using lambda, wherein lambda is a rational number and wherein lambda is an adaptively adjusted for each video unit of the video and performing a conversion between the current video unit and a bitstream of the video.

Abstract: A method for coding image information includes generating prediction information by predicting information on a current coding unit, and determining whether the information on the current coding unit is the same as the prediction information. When the information on the current coding unit is the same as the prediction information, a flag indicating that the information on the current coding unit is the same as the prediction information is coded and transmitted. When the information on the current coding unit is not the same as the prediction information, a flag indicating that the information on the current coding unit is not the same as the prediction information and the information on the current coding unit are coded and transmitted. At the generating of the prediction information, the prediction information is generated by using the information on the coding unit neighboring to the current coding unit.

Abstract: A digital PSF for use in a dual modulation display. The invention allows the use of less than optimal point spread (PSF) functions in the optics between the pre-modulator and primary modulator of a dual modulation projection system. This technique uses multiple halftones per frame in the pre-modulator synchronized with a modified bit sequence in the primary modulator to produce a compensation image that reduces the errors produced by the sub-optimal PSF. The invention includes the application to dual modulation and dual modulated 3D viewing systems.

Abstract: An image decoding method using artificial intelligence (AI), including obtaining feature data of a current optical flow and feature data of current differential data from a bitstream corresponding to a current image; obtaining the current optical flow by applying the feature data of the current optical flow to a neural-network-based first decoder; applying at least one of the feature data of the current optical flow and feature data of a previous optical flow to a first preprocessing neural network; obtain a first concatenation result by concatenating feature data obtained from the first preprocessing neural network with the feature data of the current differential data; obtaining the current differential data by applying the first concatenation result to a neural-network-based second decoder; and reconstructing the current image using the current differential data and a current predicted image generated from a previous reconstructed image based on the current optical flow.

Abstract: A method of reconstructing an optical flow by using artificial intelligence (AI), including obtaining, from a bitstream, feature data of a current residual optical flow for a current image; obtaining the current residual optical flow by applying the feature data of the current residual optical flow to a neural-network-based first decoder; obtaining a current predicted optical flow based on at least one of a previous optical flow, feature data of the previous optical flow, and feature data of a previous residual optical flow; and reconstructing a current optical flow based on the current residual optical flow and the current predicted optical flow.

Abstract: An image sensing device and a photographing device including the same are disclosed. The image sensing device includes a pixel array configured to have a first pixel and a second pixel that are different from each other in terms of at least one of an effective measurement distance, temporal resolution, spatial resolution, and unit power consumption, and a timing controller configured to determine whether a distance to a target object is equal to or less than a predetermined threshold distance, and selectively activate any one of the first pixel and the second pixel according to the result of determination.