Data denoising method and related device

An execution device, comprising a processor and a memory, wherein the processor is coupled to the memory;

• the memory is configured to store a program, wherein the program, when executed by the processor, instructs the processor to perform the following operations:
• obtaining first prediction information, wherein the first prediction information indicates predicted noise between second noisy data and the clean data, the first prediction information is generated by a first feature processing network, and the second noisy data is the first noisy data, or the second noisy data is noisy data obtained by performing at least one denoising operation on the first noisy data;
• obtaining second prediction information, wherein the second prediction information indicates a square of the predicted noise between the second noisy data and the clean data, or the second prediction information indicates a square of a predicted distance between the first prediction information and actual noise, the actual noise comprises actual noise between the second noisy data and the clean data, the second prediction information is generated by a second feature processing network, and both feature information processed by the first feature processing network and feature information processed by the second feature processing network are obtained based on the second noisy data;
• generating, based on the first prediction information and the second prediction information, distribution information corresponding to the target denoising operation;
• and
• performing sampling in distribution space to which the distribution information points, to obtain denoised data corresponding to the second noisy data.

A computer program product, wherein the computer program product comprises a program;

• and the program, when run on a computer, instructs the computer to perform following operations:
• obtaining first prediction information, wherein the first prediction information indicates predicted noise between second noisy data and the clean data, the first prediction information is generated by a first feature processing network, and the second noisy data is the first noisy data, or the second noisy data is noisy data obtained by performing at least one denoising operation on the first noisy data;
• obtaining second prediction information, wherein the second prediction information indicates a square of the predicted noise between the second noisy data and the clean data, or the second prediction information indicates a square of a predicted distance between the first prediction information and actual noise, the actual noise comprises actual noise between the second noisy data and the clean data, the second prediction information is generated by a second feature processing network, and both feature information processed by the first feature processing network and feature information processed by the second feature processing network are obtained based on the second noisy data;
• generating, based on the first prediction information and the second prediction information, distribution information corresponding to the target denoising operation;
• and
• performing sampling in distribution space to which the distribution information points, to obtain denoised data corresponding to the second noisy data.