A System for Video Recommendation using Visual Saliency, Crowdsourced and Automatic Annotations

We’ll present at ACM Multimedia 2015 in the Demos Track a system for content-based video recommendation that exploits visual saliency to better represent video features and content.

The system is demonstrated in a Social Network we implemented. Visual saliency is used to select relevant frames to be presented in a web-based interface to tag and annotate video frames in the social network; it is also employed to summarize video content to create amore effective video representation used in the recommender system. The system also exploits automatic annotations from CNN-based classifiers on salient frames and user generated annotations.

In the demo article we evaluate several baseline approaches and show how the proposed method improves over them.

The system started as a project developed by the student Saverio Meucci as final work for the Bachelor of Art at the Faculty of Engineering of University of Florence under the supervision of prof. Alberto Del Bimbo, and his assistants Andrea Ferracani and Daniele Pezzatini.

The work was born from an idea by Andrea Ferracani and Daniele Pezzatini, researchers at MICC and prof. Alberto Del Bimbo collaborators.

About Andrea Ferracani

I'm an interaction designer and an intelligent web applications developer. My research interests focus on machine learning, collective intelligence, rich internet applications, social networks analysis and the semantic web.

2 Responses to A System for Video Recommendation using Visual Saliency, Crowdsourced and Automatic Annotations

  1. Pingback: MICC - Media Integration and Communication Center

  2. Pingback: MICC will be at ACM Multimedia 2015 | Media Integration and Communication Center

Leave a Reply

Your email address will not be published. Required fields are marked *