CalligraphyNet

A quick implementation of CNN to recognize Chinese characters in different calligraphy styles.

Please refer to this page for problem description and download raw data .

• preprocessing.py converts raw images into standardized images 224x224 on a white background.
• data_loader.py loads data from folder and splits into training and validation dataset.
• trainer.py builds and trains a CNN model.