Source: ACM SIGGRAPH Citation
ACM SIGGRAPH recognizes Noah Snavely with the 2014 Significant New Researcher Award for his high impact work on synthesizing 3D models from internet scale photo collections.
Snavely’s breakthrough paper in SIGGRAPH 2006, “Photo Tourism: Exploring Photo Collections in 3D”, opened a new area of research at the boundary of graphics and vision. This work developed methods for navigating, browsing, and reconstructing models of the world from community photo collections. With the explosion of consumer cameras and photo-sharing sites, community photo collections provide a wealth of visual information of locations from all over the world, collected and shared by photographers using various cameras, captured at different instants of time and over long time periods, and in differing weather conditions. Snavely’s thesis introduced a structure from motion algorithm to reconstruct 3D scenes from such unstructured photos collections of millions of images from the Web. This research has spawned research sub-areas in graphics and vision on harnessing Internet scale photo collections for better models and understanding of the world.
Snavely’s subsequent paper “Finding Paths through the World’s Photos” in SIGGRAPH 2008, presented a powerful image-based approach to navigating photo collections. In follow up work Snavely also showed how this approach to 3D reconstruction from photo collections can be scaled to reconstructing entire cities (“Building Rome in a Day” ICCV 2009). Snavely’s research continues to explore the rich information available in photo collections including recognizing locations through images, developing games to collect images for more accurate 3D reconstruction, and combining graphics models of illumination with statistical analysis of photo collections to recover shading and visibility information of real-world scenes, among others.
Snavely’s research draws on techniques from a broad range of areas including graphics, computer vision, graph algorithms, and social network analysis. His work has had a tremendous impact on the graphics and vision research communities, and consumers in the real world. Products developed based on this research have been used by millions of users. His publicly released software, Bundler, has been widely used by researchers, practitioners and consumers in many application domains to reconstruct models of scenes from unstructured photo collections. The impact of Noah’s work extends far beyond computer graphics and vision and has been used in far reaching application domains like archaeology.
We are pleased to recognize his vision with the 2014 Significant New Researcher Award.