Skip to content

I wanted to build my own implementation of waifu2x using Keras and TensorFlow, but I ended up using a slightly different architecture.

License

Notifications You must be signed in to change notification settings

Exusai/SuperResolutionWaifuNN

Repository files navigation

SuperResolutionWaifuNN

PortfolioSRWNNVid.mp4

I wanted to build my own implementation of waifu2x using Keras and TensorFlow, but I ended up using a slightly different architecture.

Image dataset colected using PixivUtil2.

Flutter implemetation using tflite srwnn_mobile.

Training

The model was trained on about a thousand images. alt text Better quality of the gif: https://youtu.be/GCOe5kMTM5E

Test (using 8k+ images for training)

Input/Output

alt text

Tests 1

The first image is the low ressolution input, the second is the generated image and the third is the original high resolution image. alt text

Tests 2

alt text

Denoise test

alt text

Deblur test

alt text

Test (using 1.5k images for training)

Tests

The first image is the low ressolution input, the second is the generated image and the third is the original high resolution image.

Test 1

alt text

Test 2

alt text

Test 3

alt text

Test 4

alt text

Test 5

alt text

Test 6

alt text

Test 7

alt text

Test using non anime picture

alt text

About

I wanted to build my own implementation of waifu2x using Keras and TensorFlow, but I ended up using a slightly different architecture.

Topics

Resources

License

Stars

Watchers

Forks