Skip to content
Analyzing paintings to find out their year of creation
Jupyter Notebook
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Paper
visuals
.gitignore
Classifying_Paintings_by_Art_Epoch_with_CNN.ipynb
LICENSE
README.md

README.md

Classical Paintings Forensics

Analyzing paintings to find out their year of creation. By using CNNs and combined with a pre-trained VGG19 model we can find out in what year the painting was painted. In this research we are distributing a dataset of color paintings into multiple classes based on which art epoch they were painted in.

This project is initially inspired by the Kaggle task to detect the author of a given painting: https://www.kaggle.com/c/painter-by-numbers

This is an on-going research, the results reflected are preliminary.

The dataset is not included, you can generate your own dataset either from the kaggle task linked above or directly from www.wikiart.org

Current results

alt text alt text

More information and analysis of current results in the attached paper
You can’t perform that action at this time.