GHio-Ca: An Android Application for Automatic Image Classification

• Published in: Smart Objects and Technologies for Social Good Vol. 233; pp. 248 - 257
• Main Authors: Polonio, Davide, Tavella, Federico, Zanella, Marco, Bujari, Armir
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2018; Springer International Publishing
• Series: Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering
• Subjects: Android; Computer vision; Image recognition; Online social networks; Social media sensing
• ISBN: 9783319761107; 3319761102
• ISSN: 1867-8211; 1867-822X
• DOI: 10.1007/978-3-319-76111-4_25

Online social networks (OSN) have revolutionized many aspects of our daily lives and have become the predominant platform where content is consumed and produced. This trend coupled with recent advances in the field of Artificial Intelligence (AI) have paved the way to many interesting features, enriching user experience in these social platforms. Photo sharing and tagging is an important activity contributing to the social media data ecosystem. These data once labeled constitute a fruitful input for the system which is exploited to better the services of interest to the user. However, these labeling activity is imperfect and user subjective, hence prone to errors inherent to the process. In this paper, we present the design and the analysis of an Android app (namely GHio-Ca), an automatic photo tagging service relying on state-of-the-art image recognition APIs. The application is presented to the user as a camera app used to share pictures on social networks while relying on external services to automatically retrieve tags best representing the picture theme. Along with the system description we present a user evaluation involving 30 subjects.