Lightweight, wearable cameras allow collecting vast quantities of video data that approximate the visual field of the camera wearer. Scientists are mounting these devices on people in order to analyze the visual experiences of infants, for example. Egocentric cameras have also been used to study human and animal cognition, human-human social interaction, human-robot interaction, and human expertise in complex tasks. Meanwhile, from a computer science perspective, these devices record huge volumes of imagery that require new automatic techniques for effective browsing, search, and visualization. Initial technical advances have included personalized video summarization, understanding of social saliency, activity analysis, human interaction recognition, and focus of attention modeling. As these techniques become more advanced, we expect wearable cameras to unlock a variety of consumer applications, including helping people suffering from memory loss recall to assisting people with visual impairments.

We believe this is only the tip of the iceberg, and that egocentric cameras and computer vision could transform how scientists conduct their research by collecting and analyzing high-density, high-volume video data. There is a long history of connections and collaborations between human and computer vision; the pioneering early work of Dr. David Marr made immense contributions to both fields, for example. Since then, each field has progressed tremendously and has grown into its own large research area with its own subareas and topics. The emerging area of egocentric vision provides a unique opportunity to strengthen the connection between human and computer vision.

This workshop brings together leading scholars from human and computer vision, to explore opportunities for collaboration and to help plan the future of this emerging area. The specific goals are:

The workshop is supported in part by a grant program in memory of Vincent and Elinor Ostrom, both of whom served on the faculty of IU for many years. Elinor Ostrom was awarded the Nobel prize in economics in 2009. Moreover, this workshop is a kick-off event for a broader $2.5 million investment by IU's Emerging Areas of Research Program into investigating the connections and collaborations between human and machine learning.