Materials in the wild

Our daily lives bring us in contact with a rich range of materials that contribute to both the utility and aesthetics of our environment. Human beings are remarkably good at perceiving subtle distinctions in material appearance (e.g., is this fabric silk or cotton? is this surface granite or laminate?).
In my group, we are working on understanding how humans perceive materials, and using this knowledge to drive better graphics and vision algorithms.
To make progress on understanding and modeling materials, we need large-scale data about real-world materials. I will describe OpenSurfaces, a crowd-sourced database of thousands of material examples segmented from consumer photographs. I will show various uses of such large-scale material data in material recognition, intrinsic image decomposition, material-based image browsing and design. This work has applications in many domains: in virtual and augmented reality, in e-commerce and retail, and in industrial and interior design.