330,000,000 images and counting
Welcome to AMOS, the Archive of Many Scenes!
AMOS is a collection of long-term timelapse imagery from publicly accessible outdoor webcams around the world. We explore how to use these images to learn about the world around us, with a focus on understanding changes in natural environments and understanding how people use public spaces.
To support these applications, we work on fundamental research in camera geolocation, camera calibration, camera registration to GIS data, and the automatic annotation of events and objects in a scene.
We encourage you to learn more about the AMOS dataset, project participants, and publications. Options for browsing the dataset and contributing webcams to the archive are available through the links on the right.
This project gratefully acknowledges a wide variety of funding sources, including initial funded through NSF (RI-IIS 0546383), "CAREER: Passive Vision, What Can Be Learned by a Stationary Observer". Current work on the project is supported through:
- NSF (CGV-1111398), "Analyzing Images Through Time", project page (link),
- NSF (EF-1065734), "Continental-scale Monitoring, Modeling and Forecasting of Phenological Responses to Climate Change", and
- DARPA grant (D11AP00255) "ContextualEyes: A Context-Aware Surveillance System", (to Nathan Jacobs).
- Washington University I-CARES program, "Re-Purposing the Global Webcam Network as an Atmospheric/ Environmental Observation. Sensor"
This project also gratefully acknowledges AWS Convergence Technologies Inc. for allowing us to archive a collection of images from the weatherbug camera network.