Multimedia Retrieval project

For the master course Multimedia Retrieval, we had to retrieve items from a dataset using another item and different similarity measures.

The first step was to normalize the models in the dataset, to orient them and fit them in a unit box. The models are visualized in different modes using OpenGl. The user can switch between 4 different similarity measures to get the ten most similar objects in the dataset.

For more information, I refer the reader to the project report. Below is one table from the report, that shows the different similarity measures and the five closest items the measures have found for the teddybear in the top left.