MediaGLOW

Photo organizer based on multi-faceted photo similarity

MediaGLOW is an interactive visual workspace designed to address the growing number and size of digital photo libraries.

It uses attributes such as visual appearance, GPS locations, user-assigned tags, and dates to filter and group photos. An automatic layout algorithm positions photos with similar attributes near each other to allow users to serendipitously find multiple relevant photos.

To let users make use of similarity layouts, we created an interactive visual workspace called MediaGLOW that presents a photo collection based on different similarity criteria. We currently offer four different similarity criteria: temporal, geographic, tag, visual. Temporal similarity is computed from the difference between photo creation times. Geographic similarity is based on the distance between latitude-longitude pairs. Tag similarity is computed using the Jaccard similarity coefficient of tags shared across photos. Our visual similarity is determined by an image classifier trained on manually tagged photos that compares predicted likelihoods for tags. In addition to grouping photos by similarity, MediaGLOW also provides three filters that restrict the time range, the geographic location, and the tags assigned to matching photos.

MediaGLOW integrates a variety of visualization and interaction techniques with different similarity criteria, enabling users to find relevant photos by proximity and by attribute filters. For placing photos in the 2D workspace, we chose a graph layout mechanism that visually indicates similarity among photos in the space while optimizing desired distances between photos. While grid-based layouts are more common for photo applications, they cannot accurately present similarity by proximity. Furthermore, while some similarity criteria, such as time, may naturally be visualized in one dimension, multi-dimensional similarity criteria can be visualized better in a two-dimensional layout.

Technical Contact

Related Publications

2010
Publication Details
  • JCDL 2010
  • Jun 21, 2010

Abstract

Close
Photo libraries are growing in quantity and size, requiring better support for locating desired photographs. MediaGLOW is an interactive visual workspace designed to address this concern. It uses attributes such as visual appearance, GPS locations, user-assigned tags, and dates to filter and group photos. An automatic layout algorithm positions photos with similar attributes near each other to support users in serendipitously finding multiple relevant photos. In addition, the system can explicitly select photos similar to specified photos. We conducted a user evaluation to determine the benefit provided by similarity layout and the relative advantages offered by the different layout similarity criteria and attribute filters. Study participants had to locate photos matching probe statements. In some tasks, participants were restricted to a single layout similarity criterion and filter option. Participants used multiple attributes to filter photos. Layout by similarity without additional filters turned out to be one of the most used strategies and was especially beneficial for geographical similarity. Lastly, the relative appropriateness of the single similarity criterion to the probe significantly affected retrieval performance.
2009
Publication Details
  • IUI '09
  • Feb 8, 2009

Abstract

Close
We designed an interactive visual workspace, MediaGLOW, that supports users in organizing personal and shared photo collections. The system interactively places photos with a spring layout algorithm using similarity measures based on visual, temporal, and geographic features. These similarity measures are also used for the retrieval of additional photos. Unlike traditional spring-based algorithms, our approach provides users with several means to adapt the layout to their tasks. Users can group photos in stacks that in turn attract neighborhoods of similar photos. Neighborhoods partition the workspace by severing connections outside the neighborhood. By placing photos into the same stack, users can express a desired organization that the system can use to learn a neighborhood-specific combination of distances.