240 likes | 255 Views
PhotoCat is an innovative application that helps users manage and share their digital photo collections. Powered by time, location, and co-presence metadata, it provides a new and convenient way to organize and retrieve photos.
E N D
An online photo organization and retrieval system www.photocat.org Is213-photocat@sims.berkeley.edu Carrie Burgener carrie@sims Scott Fisher sbfisher@sims Andrea Nelson andrea@sims Mike Wooldridge mikew@sims
Presentation Overview • Project Goals • Personas and Scenarios • Design Rationale • Interactive Prototype • Lessons Learned
Why Build the System? • People tend to take lots of digital photos. Digital photos cost practically nothing to shoot and cameras can hold hundreds of them. • It’s inconvenient to classify large numbers of photos either at the time of capture or after you’ve downloaded them to your PC. • Better organization can help people share photos, and sharing is an important reason people take digital photos in the first place.
Project Goals • To create an application that makes use of time, location, and co-presence metadata to • help users manage their digital photo collections • help users share their digital photos • To develop an interface that presents photo collections in a new and innovative way. • Focus on camera-phone photos.
Point-of-Capture Metadata • Time • “9:30 a.m., March 31, 2005” • Location • “Cell ID 1742” or “GPS Coordinates 37.98363, -122.04845” (Can be translated into a informative description: “Berkeley, CA” ) • Co-Presence • Bluetooth ID 000db501b509 (Can be translated into a informative description: “Mike Wooldridge” )
Initial Hunches The system should: • Automatically organize photos for users • Be useful for large and small collections of photos • Allow pictures to be filtered based on -- Date, time, and day of week -- Location -- People nearby
Preliminary Interviews • Demographics • Six camera phone users between the ages of 25 and 44 • Three of the interviewees are married, and one has children in the home • Photo-Taking Habits • Personal Photo Use • Social Photo Use • Photo Organization Strategies
Preliminary Interviews • Demographics • Photo-Taking Habits • Took pictures they wouldn’t have before • Less than 5 per month to more than 150 photos taken per month • Personal Photo Use • Social Photo Use • Photo Organization Strategies
Preliminary Interviews • Demographics • Photo-Taking Habits • Personal Photo Use • Two interviewees reported liking to go back through their camera phone pictures • Social Photo Use • Photo Organization Strategies
Preliminary Interviews • Demographics • Photo-Taking Habits • Personal Photo Use • Social Photo Use • Share interesting moments with relatives or friends who were far away • Share with a person who was in the photograph • Photo Organization Strategies
Preliminary Interviews • Demographics • Photo-Taking Habits • Personal Photo Use • Social Photo Use • Photo Organization Strategies • The most basic organization strategy was to dump all pictures into a single folder. • Two common methods are to store photos in folders based on location, date, or event. • Two of the interviewees rename all of their photo files to describe who or what is in the picture
Personas Steve Steep : Adventure Recorder Takes pictures of events and gatherings. Shares primarily with people who were present or who are in his social network. Has use for place-based metadata. Uses photos to mark personal accomplishments.
Personas Sarah Jones: Distance Communicator Takes a lot of pictures of her daily life and shares a select group of friends. Pictures are stored in a semi-organized way, and new pictures are shared shortly after time of capture). Has use for time-based and place-based metadata.
Personas Jake Parker : Event Cataloger Takes pictures as part of his social role. Shares the majority of pictures taken. Almost all of the pictures are from social gatherings, and have one or more friends in the frame. Has use for time-based, event-based, and co-present metadata.
Primary Persona Darla Garcia: Family Historian Takes pictures of family/personal events. Only shares with a tight knit social group, if at all. Wants an organizational system that will facilitate finding pictures from a large selection. Has use for person-based and time-based metadata.
Design Priorities • Design around browsing by thumbnails • Reflect the ranking of importance of metadata for our interviewees: 1. Date & Time 2. Place 3. Co-present persons • Ultimately, create a visual system that combines metadata & thumbnails • Method for collecting photos of interest
Key Points • Drop Downs • act as filters • When was it? Who Was there? Where Was it? • Location represented as city and state • Bullseye can be manipulated in addition to drop downs • Photo bin allows access to primary actions on main screen • “Album” function allows for additional annotation, cataloging by subjective notion of “event” Bullseye Filter Menus Photo Bin
Paper Prototype Lessons 1. There needs to be a stronger mapping between the filters and the bullseye view. 2. Users did not understand the mapping between the "Actions" section of the screen and the Photo Bin. 3. Participants thought that the petals signified who the picture was shared with 4. All of the users expected albums to show up in a separate album view. 5. Users assumed that some elements of the bullseye view were clickable, and that clicking them would allow them to filter the photos in the collection. 5 1 2 3 4
Implementation Rationale • HTML/DreamWeaver is quick and known by all group members • ASP server-side technology chosen by default—2 group members familiar with VB. • Access allows quick DB creation, query generation & easy import/export from Office. • CSS Z-Order and absolute positioning needed for dynamic drawing/filtering. • JavaScript for mouse over details. • http://fusion.sims.berkeley.edu/sbfisher/index.asp
Additional Lessons Learned • Screen space is not infinite. Even with small thumbnails, displays have limits. • You can’t aim to please everyone. Our current system seems to work well for time and cluster overview for pictures—which allows people to see “event” groupings visually. It works less well for instant detailed photo previewing because of small thumbnail size. • Accurately simulating filtering/finding entails having a reasonable-sized data set and makes “wizard of oz” or mock-up techniques difficult. • Creating true interactivity is time consuming.
Thank you! For more information: www.photocat.org