Multimodal Alignment for Educational Resources by Machine Learning

This ongoing project aims at being a stepping-stone towards the building of high quality enriched open courseware. The first goal of this project is to build algorithms and tools allowing for the production of multimodal alignments between courseware of diverse nature. These alignments will be used (as the second goal of this project) to offer new possibilities to navigate in an intermodal fashion from video to text, text to video, with text being possibly typewritten, handwritten, or just slides.

Demos

Synchronized video
Work interface

Documentation (french)

Development section