Shypd.aiThe largest open index of AI toolsdcscn-super-resolution is an open-source AI tool that provides a TensorFlow implementation for Single-Image Super-Resolution (SISR). It uses deep CNNs with skip connections and network in network to enhance image resolution.
Trending
About
dcscn-super-resolution is a TensorFlow implementation of a deep learning-based Single-Image Super-Resolution (SISR) model, named DCSCN. It is based on the paper "Fast and Accurate Image Super Resolution by Deep CNN with Skip Connection and Network in Network." The model utilizes Deep CNNs with Residual Net, Skip Connection, and Network in Network for feature extraction, and parallelized 1x1 CNNs for image reconstruction. It also incorporates features like Pixel Shuffler, Transposed-CNN (optional), Self Ensemble, Clipping Normalization, and Dynamically loaded training images. The tool aims to provide a lighter yet effective solution for image super-resolution, with options for both larger, performance-oriented models and smaller, more compact versions.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
