|
|
|
Media Presence
Overview
We are investigating personal media management and enhancement. Within five years we expect terabyte PCs that will store everything including audio clips, documents, email, voice-mail, images, and videos. This deluge of media poses two challenges: (1) management; and (2) improving the value of media by enhancements. Media management includes organization, extracting meaning, indexing, search, navigation, and distribution. Enhancements include ensuring quality capture, post-processing, editing (e.g., composing stories), indexing, and annotating the media to make it more meaningful (e.g. adding captions to photos). Our research will cover creation, transmission, navigation, presentation, and annotation. We will not investigate indexing and speech/image/video analysis (e.g. speech to text), but will collaborate with others for such technologies. We are also interested in making media capture easier. One challenge is to encourage people to take their personal memorabilia out of the shoebox and store them online. Completeness is critical. An underlying technical challenge is ensuring that this information will be readable by future devices for indefinite preservation to compete with paper.
The MyLifeBits project is concerned with putting all personal media online (where media includes documents, in addition to images, audio and video). The first phase was devoted to capturing most of Gordon Bells information: about 20,000 documents (scanned/OCRed), 40,000 emails, 8,000 photos, 7 GB of music, and 3 GB of video. The next phase of MyLifeBits is the creation of a personal media database. All media may be annotated with text or audio, and assigned to sets/categories. Query results from the database will be displayed in several ways, including the familiar displays of Windows Explorer. They may also be rendered in a timeline. Selected results may be gang-annotated. Or, they may be dragged and dropped into authored stories (time-lines, photo albums, or highlight reels). MyLifeBits will use the full text search supported in SQL server. We were previously known as the telepresence group. When we began research in late 1995, it was clear that the killer app for telepresence was telepresentations. We produced the online version of the ACM '97 conference in a number of live and on-demand formats. We also worked on enabling network multicast capabilities for PowerPoint. The Multicast PowerPoint Add-in allows a presenter to multicast PowerPoint slides, including animations and effects, to a group of viewers. This led to the development of the Fcast and ECSRM reliable multicast protocols, which are planned for inclusion in future versions of Windows (ECSRM was rolled into PGM, which shipped in WindowsXP). We also built the Mping tool for testing multicast. The GazeMaster project aimed to restore gaze-awareness and a sense of space to video-conferencing via computer vision and graphics techniques in software, with no special hardware required. People
Projects
Selected Publications
A Call For The Home Media Network, Microsoft Research Technical Report, MSR-TR-2001-52, May 2001. Gemmell, Jim, Bell, Gordon, Lueder, Roger, Drucker, Steven, and Wong, Curtis, MyLifeBits: Fulfilling the Memex Vision, ACM Multimedia '02, December 1-6, 2002, Juan-les-Pins, France, pp. 235-238. Word (1.4 MB) PDF (297 KB) Other publications/talks available from individual researcher's home pages: Jim Gemmell; Gordon Bell Links We are part of the Bay Area Research Center in San Francisco. You can virtually wander the Halls of BARC |
||||||||
