Classifiying archaeological artefacts from the database DIME by Moesgaard Museum. README will be updated.
Commands: python3 -m venv pyvenv pip install -r requirements.txt
Idea and steps:
- Sanitize the dataset and investigate different techniques including different image processing techniques or CNN for feature extraction:
https://towardsdatascience.com/hog-histogram-of-oriented-gradients-67ecd887675f
https://medium.com/@deepanshut041/introduction-to-sift-scale-invariant-feature-transform-65d7f3a72d40
https://towardsdatascience.com/exploring-feature-extraction-with-cnns-345125cefc9a
This should be seen as a semi supervised learning approach s.t. images are clustered with simular features using either SVMs or K-NN.
Goal: To discard non artefacts or images that are bad.
The information based on the sanitized dataset:
Start classifying all the different kind of artefacts that exists in DIME, based on features and labels.
Might think using transfer learning or active learning in this case.
Train a CNN or pre-trained ViT model for this purpose.
Random Search for approx. best model.