7results about How to "Reduce bit rate" patented technology

The invention discloses an attribute information decoding and coding method, device and equipment, and belongs to the technical field of communication, and the decoding method comprises the steps: carrying out the decoding and inverse quantization of to-be-decoded data, and obtaining a reconstruction transformation coefficient corresponding to each node; the reconstruction transformation coefficients comprise a first reconstruction transformation coefficient corresponding to the brightness component, a second reconstruction residual transformation coefficient corresponding to the first chrominance component and a third reconstruction residual transformation coefficient corresponding to the second chrominance component; determining a second reconstruction transformation coefficient according to the first reconstruction transformation coefficient and the second reconstruction residual transformation coefficient; executing at least one of the following steps: determining a third reconstruction transformation coefficient corresponding to the second chrominance component according to the second reconstruction transformation coefficient and a third reconstruction residual transformation coefficient; determining a third reconstruction transformation coefficient according to the first reconstruction transformation coefficient and the third reconstruction residual transformation coefficient; and determining a third reconstruction transformation coefficient according to the first reconstruction transformation coefficient, the second reconstruction transformation coefficient and the third reconstruction residual transformation coefficient.

The invention provides a video coding method and device and a video decoding method and device, and relates to the technical field of video coding and decoding. Comprises: obtaining a student model; the student model is a model obtained by guiding a first network model to train through a teacher model, the teacher model is a model obtained by training a second network model, and when the same video is coded based on the first network model and the second network model respectively, the first network model and the second network model are selected; the code rate of the coding result of the first network model is smaller than that of the coding result of the second network model; and obtaining a coding result of the video to be coded according to the student model. According to some embodiments of the invention, in a video coding scheme for carrying out video coding by using a deep learning network model, the video reconstruction quality is ensured, and the code rate of a video coding result is reduced at the same time.

This application provides a decoding and encoding method, apparatus, and device. The method includes: obtaining a target probability distribution model corresponding to a first scaling factor based on a first bitstream corresponding to a current block; decoding a second bitstream corresponding to the current block based on the target probability distribution model corresponding to the first scaling factor to obtain corrected residual features corresponding to the current block; and determining a reconstructed image block corresponding to the current block based on the corrected residual features. This application improves encoding and decoding performance.

The invention relates to a video compression model processing method, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps that a to-be-trained student model and a trained teacher model are acquired, the teacher model and the student model are video compression models based on a neural network, the teacher model is configured to realize video compression of a first code rate, the student model is configured to realize video compression of a second code rate, and the first code rate is greater than the second code rate; video compression is carried out on the sample video through the student model and the trained teacher model, training losses are constructed based on compression results obtained in respective compression processes, and the training losses comprise distillation losses between the student model and the trained teacher model and reconstruction losses of the student model; and performing parameter adjustment on the student model according to the training loss until the model converges to obtain a trained student model. By adopting the method, it can be ensured that the compressed video has low code rate and high video quality at the same time.

The deep learning-based video multi-level adaptive compression cross-layer transmission optimization method and system of the application relates to the technical field of video transmission. By designing a compression level selection model based on deep learning, the optimal compression level is output based on the original video data and network environment parameters, and the corresponding compression strategy of the optimal compression level is obtained. A transmission strategy selection model is designed to generate a transmission strategy based on video quality parameters and network environment parameters. A global utility function is defined to calculate the global utility value when the compression strategy and the transmission strategy are assumed to be executed. An optimization objective function is designed, which inputs the compression strategy and the transmission strategy and outputs the optimized compression strategy and the transmission strategy. A condition judgment logic is set to select the final strategy based on the compression strategy and the transmission strategy before and after optimization, thereby achieving global optimization of video compression and transmission.

This application discloses an HDR video coding method, apparatus, device, and medium based on texture-luminance perceptual masking, relating to the field of video coding. The method includes: converting the linear luminance average value of the area to be encoded to the target luminance domain to obtain a first logarithmic value; determining a luminance masking component based on the first logarithmic value; performing a frequency domain transformation on the prediction residual of the area to be encoded to obtain frequency domain coefficients, accumulating the energy of the frequency domain coefficients to obtain an energy value, and converting the energy value into a texture masking component; mapping the luminance masking component and the texture masking component to the same numerical interval, and fusing the two components after unifying the interval to obtain a perceptual factor; adjusting the original Lagrange multipliers based on the perceptual factor, constructing a rate-distortion cost function using the adjusted Lagrange multipliers, and performing video coding processing on the area to be encoded according to the rate-distortion cost function to obtain the encoded HDR video. This application significantly improves coding quality and efficiency.

The embodiment of the application discloses a coefficient coding and decoding method, an encoder, a decoder and a computer storage medium, which are applied to the decoder, and the method comprises the following steps: based on the position of a to-be-decoded quantization coefficient, a first context model adaptive binary arithmetic coding mode is used to analyze a code stream to obtain the value of non-zero identification information; in the case that the value of the non-zero identification information is a first value, a second context model adaptive binary arithmetic coding mode is used to analyze the code stream to obtain the value of X preset identification information; in the case that the value of the Xth preset identification information is the first value, the target order of the exponential golomb code is determined, and the code stream is analyzed by using the exponential golomb decoding algorithm of the target order to obtain the residual value of the reconstructed quantization coefficient absolute value; and the value of the non-zero identification information, the value of the X preset identification information and the residual value of the reconstructed quantization coefficient absolute value are used to determine the reconstructed quantization coefficient absolute value corresponding to the quantization coefficient position.