M2CR - Multimodal Machine Translation and Image to Image translation, Loïc Barrault, LIUM and Joost van de Weijer, CVC-LAMP

[Loïc Barrault, LIUM] Multimodal Machine Translation: We will present the different architectures considered to process and integrate two different modalities in the framework of multimodal machine translation.

[Joost van de Weijer, CVC-LAMP] Image to Image translation: We will present multiway encoder-decoder networks between different image domains. This is applied to obtain various information like depth and semantic segmentation of the input images.