#29: MLDublin meets Jet.com- 1 min
Visual Search: how we made it possible for people to take a picture and search for similar products.
A multi-modal retrieval task in TREC-Vid 2018 where we use given text segments as queries to find related images or videos and then make up to a story.
Design conversational turn-taking models that can aid spoken dialogue systems in producing naturalistic interactions. Predicting turn-taking using LSTMs and multimodal data.