The amount of media that is being uploaded to social sites (such as Twitter, Facebook and Instagram) is providing a wealth of visual data (images and videos) augmented with additional information such as keywords, timestamps and GPS coordinates. Tapastreet 1 provides access in real-time to this visual content by harvesting social networks for visual media associated with particular locations, time and hashtags [1]. Browsing efficiently through harvested videos requires smart processing to give users a quick overview of their content in particular when using mobile platforms with limited bandwidth. This paper aims at presenting an architecture for testing several strategies for processing summaries of videos collected on social networks to tackle this issue.