cartoon-colorization

Evaluation

There are two ways to evaluate a model on the sketch and reference dataset.

evalPerReference.py evaluates the full model on the sketch database and one reference image. Its usage is as follows:
```
python evalPerReference.py --prototxt /path/to/deploy/ --caffemodel /path/to/caffemodel/
```
evalPerSketch.py evaluates the full model on one sketch image and the reference database. Its usage is as follows:
```
python evalPerSketch.py --prototxt /path/to/deploy/ --caffemodel /path/to/caffemodel/
```