An IDEAI tutorial opens the MultiMedia Modelling Conference 2019

Xavier Giro-i-Nieto reviewed the latest trends on multimodal deep learning

The 25th International Conference on MultiMedia Modelling kicked off this Tuesday in Thessaloniki (Greece) with a tutorial by Xavier Giro-i-Nieto on multimodal deep learning. This tutorial summarized the applied deep learning courses taught by TALP, VEU and GPI professors at UPC. The lectures reviewed the neural networks architectures that have become pervasive in most multimedia-related fields, such as neural machine translation, speech recognition or photo tagging. The focus on the tutorial was on multimodal applications, in which different data types are combined. So, for example, it reviewed recent works on lip reading, video re-dubbing or visual question answering. Slides are publicly available here.