This Git contains main code and tasks for the TUMO workshop "Analyzing Armenian Materials with AI", led by Chahan Vidal-Gorène (Calfa) and Baptiste Queuche (Calfa).
Goal: Localizing people in Armenian Newspapers of the 20th century. You will learn how to perform semantic segmentation and object detection in a structured document, with Mask-RCNN.
Full instructions: see week 1.
In partnership with the Fundamental Scientific Library of the National Academy of Sciences of the Republic of Armenia.
Goal: Introduction to multi-class problems with Mask-RCNN and different use cases, with heritage and street views materials.
Full instructions: see week 2.