It is a multimodal multilingual model that can analyse post of different languages and check the sentiment of the post, presence of profane, hate speech and targeted insult based on text-caption, comments, image caption and image ocr. Using these results finally we will decide whether the post is derogatory or not.
Sample PPT Link - https://docs.google.com/presentation/d/1EWyxc61tdlHnStahlwtdaSSs1GWCNxuk/edit#slide=id.g1ac38c917ca_0_10
Final PPT Link - https://docs.google.com/presentation/d/11q9DsBrN51BXP4qIc5SBPbVjtReJ9Wen/edit#slide=id.g9c364bf465bd883_7
Survey Report published in - https://trebuchet.public.springernature.app/get_content/fff5deac-391e-4e5c-8116-1d93bebd3810?utm_source=rct_congratemailt&utm_medium=email&utm_campaign=nonoa_20240215&utm_content=10.1007/s11042-024-18452-0