A method of preprocessing, prior to encoding with an external encoder, image data using a preprocessing network comprising a set of inter-connected learnable weights is provided. At the preprocessing network, image data from one or more images is received. The image data is processed using the preprocessing network to generate an output pixel representation for encoding with the external encoder. The preprocessing network is configured to take as an input encoder configuration data representing one or more configuration settings of the external encoder. The weights of the preprocessing network are dependent upon the one or more configuration settings of the external encoder.

A method of preprocessing, prior to encoding with an external encoder, image data using a preprocessing network comprising a set of inter-connected learnable weights is provided. At the preprocessing network, image data from one or more images is received. The image data is processed using the preprocessing network to generate an output pixel representation for encoding with the external encoder. The preprocessing network is configured to take as an input encoder configuration data representing one or more configuration settings of the external encoder. The weights of the preprocessing network are dependent upon the one or more configuration settings of the external encoder.

Systems and methods for error detection in video data is described. An encoding computing system can receive a video frame. The encoding computing system can encode and decode the video frame based on an encoding scheme. The encoding computing system can generate a frame error detection code for the decoded video frame based on an error detection code generation scheme. The encoding computing system can send the encoded video frame and the error detection code to a decoding computing system. The decoding computing system can decode the encoded video frame and generate a second error detection code using the code generation scheme. The decoding computing system can detect that the decoded video frame is corrupted by comparing the error detection code and the second error detection code.

The video sending method includes: acquiring a video stream to be transmitted; generating consecutive frame groups from the video stream, wherein setting a first frame in the current frame group to be a long-term reference frame that uses a first frame in a previous frame group as a reference during generation of at least one of the current frame group, wherein the long-term reference frame is a predictive coded frame configured to transmit a difference and a motion vector obtained by performing a comparison against the first frame in the previous frame group, the current frame group is a frame group other than the first frame group; and sending the frame groups to a receiving terminal. The video receiving method comprises: receiving the frame groups sent by a sending terminal; and restoring the frame groups to obtain the transmitted video stream.

In various examples, a frame may be encoded as multiple sub-frames. For example, data particularly relevant to conveying visual motion between frames may be encoded in a first sub-frame(s) with remaining data being encoded in a second sub-frame(s). Other information may be included in the first sub-frame(s), such as high entropy data. The high entropy data may be estimated using quantization and dequantization of macroblocks. Packet pacing may be applied at least between the encoded sub-frames. As the first sub-frame(s) may include the most important information for frame updates at the client device, if the second sub-frame(s) is not received and/or displayed the first sub-frame may be displayed providing high quality results. More error correction may be used for the first sub-frame than the second sub-frame to increase the likelihood that the first sub-frame is received at a client device.

In various examples, a frame may be encoded as multiple sub-frames. For example, data particularly relevant to conveying visual motion between frames may be encoded in a first sub-frame(s) with remaining data being encoded in a second sub-frame(s). Other information may be included in the first sub-frame(s), such as high entropy data. The high entropy data may be estimated using quantization and dequantization of macroblocks. Packet pacing may be applied at least between the encoded sub-frames. As the first sub-frame(s) may include the most important information for frame updates at the client device, if the second sub-frame(s) is not received and/or displayed the first sub-frame may be displayed providing high quality results. More error correction may be used for the first sub-frame than the second sub-frame to increase the likelihood that the first sub-frame is received at a client device.

A method of processing image data at a server is provided. Image data from one or more images is received at a preprocessing network comprising a set of inter-connected learnable weights, the weights being dependent upon one or more display settings of a display device. The image data is processed using the preprocessing network to generate a plurality of output pixel representations corresponding to different display settings of the display device. The plurality of output pixel representations are encoded to generate a plurality of encoded bitstreams. At least one selected bitstream is transmitted from the server to the display device, wherein the at least one encoded bitstream is selected on the basis of the one or more display settings of the display device.

A method of processing image data at a server is provided. Image data from one or more images is received at a preprocessing network comprising a set of inter-connected learnable weights, the weights being dependent upon one or more display settings of a display device. The image data is processed using the preprocessing network to generate a plurality of output pixel representations corresponding to different display settings of the display device. The plurality of output pixel representations are encoded to generate a plurality of encoded bitstreams. At least one selected bitstream is transmitted from the server to the display device, wherein the at least one encoded bitstream is selected on the basis of the one or more display settings of the display device.

Provided is in-loop filtering technology using a trained deep neural network (DNN) filter model. An image decoding method according to an embodiment includes receiving a bitstream of an encoded image, generating reconstructed data by reconstructing the encoded image, obtaining information about a content type of the encoded image from the bitstream, determining a deep neural network (DNN) filter model trained to perform in-loop filtering by using at least one computer, based on the information about the content type, and performing the in-loop filtering by applying the reconstructed data to the determined DNN filter model.

Provided is in-loop filtering technology using a trained deep neural network (DNN) filter model. An image decoding method according to an embodiment includes receiving a bitstream of an encoded image, generating reconstructed data by reconstructing the encoded image, obtaining information about a content type of the encoded image from the bitstream, determining a deep neural network (DNN) filter model trained to perform in-loop filtering by using at least one computer, based on the information about the content type, and performing the in-loop filtering by applying the reconstructed data to the determined DNN filter model.