Community Photo Collections

CPC example
Sample images of Notre Dame Cathedral from Flickr (from a collection of 80,000 as of Sept 2007).  Such collections are generated by a large community of photographers and hosted on Internet sharing sites.

With the recent rise in popularity of Internet photo sharing sites like Flickr and Google Images, community photo collections (CPCs) have emerged as a powerful new type of image dataset for computer vision and computer graphics research.  With billions of such photos now online, these collections should enable huge opportunities in 3D reconstruction, visualization, image-based rendering, recognition, and other research areas.  The challenge is that these collections have extreme variability, having been taken by numerous photographers from myriad viewpoints with varying lighting and appearance, and often with significant occlusions and clutter.  Our research seeks to develop robust algorithms that operate successfully on such image sets to solve problems in computer vision and computer graphics.

Team

The following people are part of the CPC team:

Overview Talk

The Google Tech Talk "Navigating the World's Photographs" by Steve Seitz, Noah Snavely, and Michael Goesele gives a good overview over our current work on community photo collections. View a video of the talk on Google Video or download the video in Flash Video (FLV) format (114 MB) or AVI format (141 MB).

Internet Vision Paper

The paper Scene Reconstruction and Visualization From Community Photo Collections by Noah Snavely, Ian Simon, Michael Goesele, Rick Szeliski, and Steve Seitz has been published in the August 2010 special issue of Proceedings of the IEEE on Internet Vision. This paper also gives a good overview of our work on community photo collections.

Projects and Publications

The following projects are based on community photo collections:

Photo Tourism Photo Tourism

Photo tourism: Exploring photo collections in 3D (PDF)
Noah Snavely, Steven M. Seitz, Richard Szeliski.
ACM Transactions on Graphics (SIGGRAPH Proceedings), 25(3), 2006, 835-846.

More information is available at the project page.

MVS reconstruction of the Duomo in Pisa
Multi-View Stereo for Community Photo Collections

Multi-View Stereo for Community Photo Collections (PDF)
Michael Goesele, Noah Snavely, Brian Curless, Hugues Hoppe, Steven M. Seitz
Proceedings of ICCV 2007, Rio de Janeiro, Brasil, October 14-20, 2007

More information is available at the project page.

Scene Summarization for Online Image Collections

Scene Summarization for Online Image Collections (PDF)
Ian Simon, Noah Snavely, and Steven M. Seitz.
Proceedings of ICCV 2007, Rio de Janeiro, Brasil, October 14-20, 2007.

More information is available at the project page.

Finding Paths through the World's Photos

Finding Paths through the World's Photos (PDF) (qt video)
Noah Snavely, Rahul Garg, Steven M. Seitz, and Richard Szeliski.
To appear in SIGGRAPH 2008.

Skeletal Graphs for Efficient Structure from Motion

Skeletal Sets for Efficient Structure from Motion (PDF)
Noah Snavely, Steven M. Seitz, and Richard Szeliski.
To appear in CVPR 2008.

Scene Segmentation Using the Wisdom of Crowds

Scene Segmentation Using the Wisdom of Crowds
Ian Simon and Steven M. Seitz.
Proceedings of ECCV 2008, Marseille, France, October 12-18, 2008.

Building Rome in a Day (PDF)
Sameer Agarwal, Noah Snavely, Ian Simon, Steven M. Seitz and Richard Szeliski
International Conference on Computer Vision, 2009, Kyoto, Japan.