Enabling crowdsourced live event coverage with adaptive collaborative upload strategies

User-generated content, such as short video snippets or tweets, is increasingly used in event coverage even by professional media outlets. Especially in unforeseen events, or when dealing with large crowds, these snippets provide unique perspectives on the scene. While uploading a tweet does not impose much load on the communication system, uploading live video at today's camera resolutions consumes a significant amount of resources. At the same time, only a fraction of the uploaded streams is suitable for event coverage (e.g., shakiness of the video, focus on the scene, obstructions). By identifying the set of relevant streams early, and postponing the upload of other content, the available network resources can be dedicated to the upload of the most relevant streams. In this paper, we propose a set of strategies to collaboratively upload the most relevant streams at high quality by utilizing freed resources. We argue that these strategies can be exchanged during runtime to adapt to user dynamics and network heterogeneity, and present initial findings on the performance of our system.