SidoPillai / neural-graph

An implementation of artistic style in TensorFlow.

neural-graph

An implementation of A neural algorithm of artistic style in TensorFlow.

Pre-Requisties

• TensorFlow
• NumPy
• SciPy
• Pillow

Data Files

• Pre-trained VGG network - insert the model in the parent directory or specify the path using the --network option.

Execution

python neural_graph.py --content <content file> --styles <style file> --output <output file>

Run python neural_graph.py --help to view a list of all options.

Use --checkpoint-output and --checkpoint-iterations to save checkpoint images.

Use --iterations to change the number of iterations (default 1000). For a 512×512 pixel content file, 1000 iterations take 60 seconds on a GTX 1080 Ti, 90 seconds on a Maxwell Titan X, or 60 minutes on an Intel Core i7-5930K. Using a GPU is highly recommended due to the huge speedup.

Results

Running it for 500-2000 iterations seems to produce nice results. With certain images or output sizes, you might need some hyperparameter tuning (especially --content-weight, --style-weight, and --learning-rate).

LICENSE for details.