proposing an addition to to the ontolex lemon model for value and directional enrichment

Dear all,

Natural language driven text describes facts, stories, concepts etc and
perception of all such happens mostly visually and as well other four
senses in reality; visual sense carries a lot in comparison. Adding a
visual (image and video) annotation to the ontolex lemon model seems a
great value addition, which can definitely provide solidity to natural
language grounding and computer vision and its associated fields. There
have been a lot of different efforts in natural grounding as well as CVPR
but such an standardized is still yet to happen; Imagenet is mentionable in
this regard as a is a large visual database mapped with wordnet. Image
describes objects or static entities, while videos are essential to
describe action or abstract items for its systematic grounding.

I am wondering if I can take a stride to focus on this direction as a
participant of the public-ontolex group and would very much like your
suggestion for the adoption, guidance or collaboration on this modeling.
Establishing this will open up many wings and directions to consolidate in
natural language grounding and as well CVPR use case oriented research for
sure.

Let's discuss.

Regards,
Saurav Karmakar

----------------------------
Dr. Saurav Karmakar
Postdoctoral Researcher,
Insight Centre,
Data Science Institute,
National University of Ireland Galway,
Galway, Ireland