Align before Attend: Aligning Visual and Textual Features for Multimodal Hateful Content Detection.

Eftekhar Hossain, Omar Sharif, Mohammed Moshiul Hoque, Sarah Masud Preum

Browse the full EACL paper archive.