metadata
license: apache-2.0
task_categories:
- question-answering
tags:
- semantic
- search
- video
size_categories:
- 1K<n<10K
This is a dataset containing lectures where the instructor draws under a document camera.
The dataset can be used to train multimodal representation learning approaches that can be used by Q&A agents. Questions may come in the form of a natural language query "what are the advantages of CNNs over Dense networks" and the system should respond back with the video clip / segment that contains the video explanation.