Көпмодальды көріністер
Related Papers
Image Pivoting for Learning Multilingual Multimodal Representations
AbstractIn this paper we propose a model to learn multimodal multilingual representations for matching images and sentences in different languages, with the aim of advancing multilingual versions of image search …...
Multimodal Frame Identification with Multilingual Evaluation
AbstractAn essential step in FrameNet Semantic Role Labeling is the Frame Identification (FrameId) task, which aims at disambiguating a situation around a predicate. Whilst current FrameId methods rely on textual …...
The Effects of Unimodal Representation Choices on Multimodal Learning
Multimodal representations are distributed vectors that map multiple modes of information to a single mathematical space, where distances between instances delineate their similarity. In most cases, using a single unimodal …...
Learning Language and Multimodal Privacy-Preserving Markers of Mood from Mobile Data
AbstractMental health conditions remain underdiagnosed even in countries with common access to advanced medical care. The ability to accurately and efficiently predict mood from easily collectible data has several important …...