Nvidia' GPUs powering intelligence platforms

Posted by Emory Kale
Nvidia just launched the GeoInt Accelerator, a GPU-accelerated geospatial intelligence platform to enable security analysts to find actionable insights quicker and more accurately rom vast quantities of raw data, images and video. Sounds great to us.
 
 
According to the Nvidia release on this topic:
 
The Nvidia GeoInt Accelerator platform provides defense and homeland security analysts with tools that enable faster processing of high-resolution satellite imagery, facial recognition in surveillance video, combat mission planning using geographic information system (GIS) data, and object recognition in video collected by drones.
 
It offers a complete solution consisting of an Nvidia Tesla GPU accelerated system, software applications for geospatial intelligence analysis, and advanced application development libraries.
 
"Today's intelligence analyst needs information based on imagery, video, signals intelligence, human intelligence and other sources, in a geospatial context and 'at rate,'" said Nick Buck, CEO Buck Consulting Group and former IT executive with the NRO Ground Directorate. "The Nvidia GeoInt Accelerator provides developers the opportunity to unleash the power of GPU computing to quickly combine the variety and volume of data sources at the speed needed for mission support, enabling new levels of performance with minimal investment."
 
In light of the general sensitivity of the human race to these sorts of things, we could be phlegmatic about this stuff, but honestly, it sounds pretty cool.
 
Nvidia's GPUs are already widely used in the defense industry for imaging, video, and signal processing.
 
 
The GeoInt Accelerator platform provides software libraries and tools to build out applications such as the following:
 
DigitalGlobe - Processes more than three million square kilometers of high-resolution imagery collected daily by satellites for current intelligence on points of interest.
 
GeoWeb 3D - Delivers native 3D GIS fusion -- including LIDAR remote sensing technology and full motion video -- without preprocessing.
 
Imagus - Real-time facial recognition from video surveillance
 
IntuVision Panoptes - Provides object detection and event-driven alerts by processing multiple real-time HD video streams.
 
LuciadLightspeed - Provides situational awareness for mission planning by overlaying image, radar, sensor data for line-of-sight analysis.
 
NerVve Technologies - Automatically detects objects in images and video streams. 
 
Oh, well. We're all one algorithm away from being put into pods and serving robot overloads. I keep saying it and all anyone can do is worry about the zombie apocalypse. Don't say I didn't warn you.