Visual Media Reasoning (VMR)*
Abstract
*Previously Web-Scale Information Integration The Visual Media Reasoning (VMR) program will create technologies to automate the analysis of enemy-recorded photos and videos and identify, within minutes, key information related to the content. Such identification will include the names of individuals within the image (who), the enumeration of the objects within the image and their attributes (what), and the image's geospatial location and time frame (where and when). Large data stores of enemy photos and video are available but cannot be easily leveraged by a warfighter or analyst attempting to understand a specific new image. The VMR program will enable users to gain insights rapidly through application of highly parallelized image analysis techniques that can process the imagery in massive federated image stores. VMR technology will serve as a force-multiplier by rapidly and automatically extracting tactically relevant information for the human analyst and alerting the analyst to scenes that warrant the analyst's expert attention. This effort will be funded in PE 0602702E, Project TT-13 beginning in FY 2013.
Document Details
- Document Type
- Accomplishment
- Publication Date
- Oct 01, 2013
- Source ID
- 7dce193cc85c3e6f7f353dd1b273ed1d
Related Documents
- Root: MACHINE INTELLIGENCE