Hand-Crafted and Learned Spatiotemporal Filters to Inform and Track Visual Saliency.
Khaled AboumerhiRalph Etienne-CummingsJonah SenguptaJohn RattrayPublished in: ISMSI (2021)
Keyphrases
- hand crafted
- visual saliency
- saliency map
- domain independent
- linguistic features
- deep learning
- visual attention
- wordnet
- automatically generated
- natural images
- eye movements
- semantic role labeling
- human detection
- moving objects
- higher order
- region of interest
- natural language processing
- bayesian framework
- human visual system
- supervised learning
- eye tracking
- search engine
- semantic information
- image data
- prior knowledge
- knowledge base