cheekybastard/fast-neural-style.tf

Feed-forward neural network for real-time artistic style transfer

fast-neural-style.tf

An implementation of [1][2].

Time: 0.05 seconds on NVIDIA TITAN X GPU for at resolution of 1200x630.

ModelName	Style	Samples
composition_v
feathers
la_muse
mosaic
the scream
udnie
wave
cubist

Requirement

Tensorflow(>= r0.11), Numpy, Scipy, Pillow

Usage

python render.py --input $INPUT_IMAGE --output $OUTPUT_FILE_NAME --model $MODEL_NAME --arch $GENERATOR_ARCH

Training

Coming soon.

License

MIT

Reference

[1] Johnson, Justin, Alexandre Alahi, and Li Fei-Fei. "Perceptual losses for real-time style transfer and super-resolution." arXiv preprint arXiv:1603.08155.

[2] Ulyanov, Dmitry, Andrea Vedaldi, and Victor Lempitsky. "Instance Normalization: The Missing Ingredient for Fast Stylization." arXiv preprint arXiv:1607.08022.

[3] Gatys, Leon A., Alexander S. Ecker, and Matthias Bethge. "A neural algorithm of artistic style." arXiv preprint arXiv:1508.06576.