1 paper
Benet Oriol, Jordi Luque, Ferran Diego +1
In this work, we propose an effective approach for training unique embedding representations by combining three simultaneous modalities: image and spoken and textual narratives. Th…