Image compression method and apparatus thereof
AssigneeWuhan Tcl Corporate Research Co., Ltd.
InventorsYunlei Xiao, Yangxing Liu
Life of this patent
Estimating the term…
Abstract
An image compression method includes: obtaining a target image and a target code rate corresponding to the target image; determining a first code rate parameter corresponding to the target code rate; and inputting the target image and the first code rate parameter into an image compression model, that has been trained, for processing to obtain a compressed image with the target code rate, wherein the image compression model is obtained by training an initial image compression model based on multiple code rate parameters including the first code rate parameter.
Claims18
An image compression method, comprising:
- obtaining a target image and a target code rate corresponding to the target image;
- determining a first code rate parameter corresponding to the target code rate;
- and
- inputting the target image and the first code rate parameter into an image compression model, that has been trained, for processing to obtain a compressed image with the target code rate, wherein the image compression model is obtained by training an initial image compression model based on multiple code rate parameters including the first code rate parameter;
- wherein:
- the first code rate parameter includes a first compression weight;
- the image compression model includes a first non-linear coding network and a compression network;
- and
- inputting the target image and the first code rate parameter into the image compression model that has been trained for processing to obtain the compressed image with the target code rate includes:
- inputting the target image and the first compression weight into the first non-linear coding network to perform a dummy-encoding process on a first compression weight example to obtain a code value corresponding to the first compression weight example via the first non-linear coding network, and using the code value to perform a convolution process on the target image to obtain a first feature map of the target image;
- and
- inputting the first feature map to the compression network to perform a compression process to obtain the compressed image.
An image compression device, comprising:
- a memory;
- a processor;
- and
- a computer program stored in the memory and executed by the processor to:
- obtain a target image and a target code rate corresponding to the target image;
- determine a first code rate parameter corresponding to the target code rate;
- and
- input the target image and the first code rate parameter into an image compression model that has been trained for processing to obtain a compressed image with the target code rate, wherein the image compression model is obtained by training an initial image compression model based on multiple code rate parameters including the first code rate parameter;
- wherein:
- the first code rate parameter includes a first compression weight;
- the image compression model includes a first non-linear coding network and a compression network;
- and
- inputting the target image and the first code rate parameter into the image compression model that has been trained for processing to obtain the compressed image with the target code rate includes:
- inputting the target image and the first compression weight into the first non-linear coding network to perform a dummy-encoding process on a first compression weight example to obtain a code value corresponding to the first compression weight example via the first non-linear coding network, and using the code value to perform a convolution process on the target image to obtain a first feature map of the target image;
- and
- inputting the first feature map to the compression network to perform a compression process to obtain the compressed image.
Dependencies are inferred from the claim text.