Shypd.aiThe largest open index of AI toolsIRCNN is an open-source deep learning tool for image restoration that uses a CNN denoiser prior. It is implemented in Matlab and can solve various inverse problems in low-level vision, including deblurring, inpainting, and super-resolution.
Trending
About
IRCNN is an open-source deep learning tool designed for image restoration, leveraging a Convolutional Neural Network (CNN) denoiser prior. It addresses various inverse problems in low-level vision, such as image deblurring, inpainting, single image super-resolution, and color image demosaicking. The tool integrates fast and effective CNN denoisers into model-based optimization methods, offering flexibility and good performance without requiring additional training for different tasks. It is implemented in Matlab, with a PyTorch version also available, and utilizes techniques like Half-Quadratic Splitting (HQS) for efficient processing. This makes IRCNN a valuable resource for researchers and developers working on advanced image processing applications.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
