Behind the Machine's Gaze: Neural Networks with Biologically-inspired Constraints Exhibit Human-like Visual Attention.
Leo SchwinnDoina PrecupBjörn M. EskofierDario ZancaPublished in: Trans. Mach. Learn. Res. (2022)
Keyphrases
- visual attention
- biologically inspired
- visual attention model
- neural network
- eye tracking
- eye movements
- saliency map
- visual search
- learning rules
- center surround
- human behavior
- spiking neural networks
- focus of attention
- natural scenes
- receptive fields
- biologically plausible
- vision system
- visual perception
- higher level
- eye tracking data
- eye tracking technology
- back propagation
- visual field
- object based visual attention
- saliency model
- visual saliency
- salient regions
- humanoid robot
- image quality
- artificial neural networks
- high level
- eye tracker
- feed forward
- multi modal
- low level