A visual embedding for the unsupervised extraction of abstract semantics.
Dario Garcia-GasullaEduard AyguadéJesús LabartaJavier BéjarUlises CortésToyotaro SuzumuraR. ChenPublished in: Cogn. Syst. Res. (2017)
Keyphrases
- high level
- low level
- visual information
- formal semantics
- automatic extraction
- visual perception
- lexical semantics
- vector space
- information extraction
- visual cues
- supervised learning
- unsupervised learning
- logic programming
- visual features
- higher level
- semi supervised
- propositional dynamic logic
- high level semantics
- semantically relevant
- learning algorithm
- visual data
- data driven
- pairwise
- metadata