Method for accelerating GPU directed at deep learning super-resolution technology

Abstract

The invention discloses a method for accelerating a GPU directed at a deep learning super-resolution technology. The method conducts concurrent processing on all the steps of a super-resolution technology which is based on deep learning and a convolutional neural network, and operates on a GPU. The concurrent processing refers to conducting concurrent task dividing on convolutions of the super-resolution technology which is based on deep learning and the convolutional neural network into millions of micro-tasks which are irrelevant to one another and can be concurrently executed in any order so as to fully exhibit the super-strong calculating capability of the GPU. Further, the method uses features of a GPU memory to cache convolution nuclear data and input image data to a shared memory and a register so as to substantially optimize calculating speeds of the convolutions. The method integrates the convolutions and non-linear layers. The method selects the optimal method for the sizes of different convolutions. According to the invention, the method accelerates the high quality super-resolution method to meet velocity requirements for processing videos, and does not cause any image quality loss.

Description

technical field [0001] The present invention relates to an image super-resolution field and a GPU acceleration method, in particular to a GPU acceleration method for deep learning super-resolution technology. Background technique [0002] Image super-resolution is to convert a low-resolution image into a high-resolution image, which has a wide range of applications in image post-processing and video nonlinear editing. Early super-resolution methods (such as bicubic) are often based on simple interpolation, which can work quickly and reliably, and are easy to integrate on chips; however, the high-resolution images obtained by these methods are of poor quality and produce significant artifacts, such as rings, Effects such as aliasing, blurring, etc. Super-resolution methods of such quality are difficult to meet the current high-quality video demands. The current super-resolution methods with advanced performance can generate high-quality images, but with huge computational o...

AI Technical Summary

Problems solved by technology

However, this method relies on a huge amount of computing overhead, and it takes 300 seconds to execute this method with a CPU for each frame (1920*1080 to 3840*2160, single channel, all tests below are based on this resolution), even if GEMM-based GPU convolution and acceleration methods also require close to 1 second for each frame, which cannot meet the needs of practical applications

Images