Showing 1–1 of 1 results for author: de Juan, A V

SneakPeek: Interest Mining of Images based on User Interaction

Authors: Daniyal Shahrokhian, Alejandro Vera de Juan

Abstract: Nowadays, eye tracking is the most used technology to detect areas of interest. This kind of technology requires specialized equipment recording user's eyes. In this paper, we propose SneakPeek, a different approach to detect areas of interest on images displayed in web pages based on the zooming and panning actions of the users through the image. We have validated our proposed solution with a gro… ▽ More Nowadays, eye tracking is the most used technology to detect areas of interest. This kind of technology requires specialized equipment recording user's eyes. In this paper, we propose SneakPeek, a different approach to detect areas of interest on images displayed in web pages based on the zooming and panning actions of the users through the image. We have validated our proposed solution with a group of test subjects that have performed a test in our on-line prototype. Being this the first iteration of the algorithm, we have found both good and bad results, depending on the type of image. In specific, SneakPeek works best with medium/big objects in medium/big sized images. The reason behind it is the limitation on detection when smartphone screens keep getting bigger and bigger. SneakPeek can be adapted to any website by simply adapting the controller interface for the specific case. △ Less

Submitted 10 December, 2017; originally announced December 2017.