Realcugan Ncnn Webassembly Save

Super-resolution processing for anime images based on WASM; 基于WASM的动漫图片超分辨率处理

Project README

RealCUGAN-ncnn-webassembly

中文

This project uses Web Assembly technology to run the Real-CUGAN model based on ncnn.

Real-CUGAN is an AI super resolution model for anime images, trained in a million scale anime dataset, using the same architecture as Waifu2x-CUNet. It supports 2x\3x\4x super resolving. For different enhancement strength, now 2x Real-CUGAN supports 5 model weights, 3x/4x Real-CUGAN supports 3 model weights.

The code implementation deeply refers to realcugan-ncnn-vulkan and ncnn-webassembly-nanodet.

Usage

Website: https://real-cugan.animesales.xyz/

iOS: is not currently supported.

Android: please open in a browser app.

PC/Mac/Linux: recommend using the latest version of Chrome or Firefox.

How to build

  1. Install emscripten:
git clone https://github.com/emscripten-core/emsdk.git
cd emsdk
./emsdk install 3.1.13
./emsdk activate 3.1.13

source emsdk/emsdk_env.sh # or add it to .zshrc etc.
  1. build:
git clone https://github.com/hanfengsan/realcugan-ncnn-webassembly.git
cd realcugan-ncnn-webassembly

git submodule update --init
sh build.sh

go run local_server.go

open in brower: http://localhost:8000

Open Source Agenda is not affiliated with "Realcugan Ncnn Webassembly" Project. README Source: hanFengSan/realcugan-ncnn-webassembly