Abstract
Saliency prediction is a very important and challenging task within the computer vision community. Many models exist that try to predict the salient regions on a scene from its RGB image values. Several new models are developed, and spectral imaging techniques may potentially overcome the limitations found when using RGB images. However, the experimental study of such models based on spectral images is difficult because of the lack of available data to work with. This article presents the first eight-channel multispectral image database of outdoor urban scenes together with their gaze data recorded using an eyetracker over several observers performing different visualization tasks. Besides, the information from this database is used to study whether the complexity of the images has an impact on the saliency maps retrieved from the observers. Results show that more complex images do not correlate with higher differences in the saliency maps obtained.
Highlights
Eight-Channel Multispectral ImageHuman observers are able to process a considerable number of visual stimuli
This study aims to cover this gap by presenting a database of multispectral images of urban scenes and their corresponding saliency maps
This article presents a database of eight-channel visible and near infrared (VIS+NIR)
Summary
Human observers are able to process a considerable number of visual stimuli. They focus their attention on certain areas of the scene or some objects that are either markedly distinct from their surroundings (e.g., mature orange fruits that hang from an orange tree) or relevant for the task they are performing (e.g., traffic lights when the subject is planning to cross a street). The subject’s visual attention mechanisms are driven by the inherent features of the stimuli (differences in shape and color from the surroundings). This is a typical instance of bottom–up visual attention [1]. Numerous models have been developed and tested for their performance in two main tasks: saliency detection (when the model’s aim is to predict and segment the most salient object in the scene [3,4,5,6]) and saliency prediction (when the model’s aim is to predict the areas to where the subject’s gaze will be directed for longer periods of time [5,7,8,9])
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have