Computational versus psychophysical bottom-up image saliency: A comparative evaluation study
| Autor: | Alexander Toet |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2011 |
| Předmět: |
Conspicuity
Visual perception Image compression Computer science Image quality Measurement procedures Content-aware Feature dimensions Adaptive content Pattern Recognition Automated Image analysis Engineering Human visual Target support Digital image storage Computer vision Image communication systems Region of interest Visualization Relative contribution Image segmentation visual search Applied Mathematics Video summarization Cognitive neuroscience of visual object recognition Psychophysical Pattern Recognition Visual Computational Theory and Mathematics Feature (computer vision) Computer Vision and Pattern Recognition Algorithms Human PCS - Perceptual and Cognitive Systems Models Neurological ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Image scaling Visual targets Image processing Fixation Ocular Artificial Intelligence Salience (neuroscience) Image quality assessment Psychophysics Humans Computer Simulation Human visual inspection Rank order Lateral interactions Visual saliency Visual search Saliency business.industry Visual effects Measurements Static and dynamic Comparative evaluations Multiscale contrast Object recognition Progressive image transmission BSS - Behavioural and Societal Sciences Artificial intelligence business Additional knowledge Software Forecasting |
| Zdroj: | IEEE Transactions on Pattern Analysis and Machine Intelligence, 11, 33, 2131-2146 |
| Popis: | The predictions of 13 computational bottom-up saliency models and a newly introduced Multiscale Contrast Conspicuity (MCC) metric are compared with human visual conspicuity measurements. The agreement between human visual conspicuity estimates and model saliency predictions is quantified through their rank order correlation. The maximum of the computational saliency value over the target support area correlates most strongly with visual conspicuity for 12 of the 13 models. A simple multiscale contrast model and the MCC metric both yield the largest correlation with human visual target conspicuity ({}0.84). Local image saliency largely determines human visual inspection and interpretation of static and dynamic scenes. Computational saliency models therefore have a wide range of important applications, like adaptive content delivery, region-of-interest- based image compression, video summarization, progressive image transmission, image segmentation, image quality assessment, object recognition, and content-aware image scaling. However, current bottom-up saliency models do not incorporate important visual effects like crowding and lateral interaction. Additional knowledge about the exact nature of the interactions between the mechanisms mediating human visual saliency is required to develop these models further. The MCC metric and its associated psychophysical saliency measurement procedure are useful tools to systematically investigate the relative contribution of different feature dimensions to overall visual target saliency. © 2011 IEEE. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|