An article detailing some of the projects being considered for the “DARPA Innovation House” should pique the interest of anyone working with video analytics and surveillance.
From the project website: “The DARPA Innovation House is a study into the feasibility of effective software design and development in a short-fuse, crucible-style living and working environment. DARPA selected imagery analysis as the topic for the effort. DARPA aims to show that small teams of highly focused, collaborative developers operating under extremely short deadlines can make breakthroughs in automatically obtaining meaning from photos, videos, geospatial data and other imagery-related data.”
The proposed areas of study include:
- Petavision: Multi-Modal Approaches to Real-Time Video Analysis
Biologically-inspired, hierarchical neural networks to detect objects of interest in streaming video by combining texture/color, shape and motion/depth cues.
- GOSE (Geospatial-Oriented Structure Extraction): Structure Extraction from Imagery
Automatic construction of a 3D wireframe of an object using as few images as possible from a variety of angles.
- Videovor: Visualization of Video Information
Software that fragments and re-models linear time of video content to cluster big pools of related data in a 3D interactive interface for human analysis.
The full list can be found here.
Several of the projects have the potential to benefit security and loss prevention systems, so their progress will be interesting to watch.