imxieyi/ waifu2x-ios
About the developer
291 Stars 
31 Forks 
MIT License 
63 Commits 
2 Opened issues 
Description
iOS Core ML implementation of waifu2x
Services available
waifu2x on iOS
Introduction
This is a Core ML implementation of waifu2x. The target of this project is to run waifu2x models right on iOS devices even without network. For macOS version please refer to waifu2x-mac.
Video support based on Metal Performance Shaders is also included in this repo. Models are loaded directly from Core ML models (see CoreML-MPS). It is meant to be run on macOS with a powerful discerete GPU through Mac Catalyst. Running it on iOS devices will significantly drop battery life and cause thermal issues. Most likely it will crash immediately.
The author is not responsible of any damage to your device.
Requirements
- XCode 9+
- iOS 11+
- macOS 10.15+ (for Mac Catalyst)
Usage
After cloning this repo, remember to update submodules:
git submodule update --init
Then open
waifu2x-ios.xcworkspace(not
waifu2x-ios.xcodeproj).
Click
Video Teston the top to pick video files. Output path will be printed in the console (starts with
Output path:).
Image format
Images with
RGBcolor space works fine. Others should be converted to
RGBbefore processing otherwise output image will be broken. Alpha channel is scaled using bicubic interpolation. Generally it runs on GPU. It automatically falls back to CPU if image is too large for Metal to process, which is extremely slow. (A bad idea)
Video format
The built-in video decoder on iOS and macOS is very limited. If your video doesn't work, you can convert to a supported format using ffmpeg:
ffmpeg -i -c:v libx264 -preset ultrafast -pix_fmt yuv420p -c:a aac -f mp4 .mp4
About models
This repository includes all the models converted from waifu2x-caffe. If you want to dig into Core ML, it is recommended that you should convert them by yourself.
You can convert pre-trained models to Core ML format and then import them to XCode. The pre-trained model can be obtained from waifu2x-caffe.
You can use the same method described in MobileNet-CoreML. You should not specify any input and output layer in python script.
A working model should have input and output like the following example:
Benchmark on images
Environment
- iPhone6s - waifu2x-ios on iPhone 6s with iOS 11.1
- iPhone8 - waifu2x-ios on iPhone 8 with iOS 11.0
- iPhone11Pro - waifu2x-ios on iPhone 11 Pro with iOS 13.1
- iPad - waifu2x-ios on iPad Pro 10.5 with iOS 11.1
-
PC - waifu2x-caffe on Windows 10 16278 with GTX 960M
### Results
All of the tests are running
denoise level 2
withscale 2x
model on anime-style images from Pixiv.
Test1
Image resolution: `600849`*
Device |
Time(s) |
|---|---|
| iPhone6s | 6.8 |
| iPhone8 | 4.0 |
| iPhone11Pro | 2.0 |
| iPad | 2.9 |
| PC | 2.1 |
Test2
Image resolution: `30003328`*
Device |
Time(s) |
|---|---|
| iPhone6s | 129.2 |
| iPhone8 | 73.5 |
| iPhone11Pro | 18.8 |
| iPad | 49.2 |
| PC | 37.5 |
Evolution
Device: iPad Image resolution: `30003328`*
Milestone |
Time(s) | RAM usage(GB) |
|---|---|---|
| Before using upconv models | 141.7 | 1.86 |
| After using upconv models | 63.6 | 1.28 |
| After adding pipeline on output | 56.8 | 1.28 |
| After adding pipeline on prediction | 49.2 | 0.38 |
| Pure MPSCNN implementation* | 29.6 | 1.06 |
*: With crop size of 384 and double command buffers.
Performance on video
About 1.78 frames per second while scaling
1080p -> 2160pon 5700XT GPU.
Runs out of memory and crashes immediately with the same video on iOS with 4GB memory.
Demo
