Using the Surrounding WEB Content of Pictures to Generate Candidates for Photo Annotation

The nowadays pervasive mobile devices embed among one or two cameras a multitude of sensors facilitating the capture of the surrounding context and thus the picture visual content is enriched with the additional metadata. Moreover, the digital storage of the pictures either on a personal device or by the use of an online photo management and sharing platform enables a series of services among which the generation of a list of individuals likely to appear in a certain picture. We propose a novel approach for annotating individuals appearing in pictures to benefit from this information gained on the one side from the sensors embedded in the devices used to capture the picture and on the other hand from the online platform hosting the pictures. Precisely, the gathered information is used for the creation of a list of candidates likely to appear in the picture that a user wants to annotate dispensing him to browse a large list of users to choose from. We make a thoroughly discussion on the opportunities brought by a collection of annotated pictures and discuss on the further expediencies that it may bring in the healthcare domain in an attempt to extend the current researchers' vision who consider that the only advantages of having an annotated collection of pictures are a better retrieval and an improved management.