Xailient’s computer vision AI is optimized for tiny devices using:
Smart Motion Detection
Many systems continuously monitor a field of view. Smart Motion Detection uses AI to look for personnel, vehicles, etc. and ignores background motion or the motion of the camera. Smart Motion Detection can trigger alerts or be used to activate systems out of a “standby” mode, and, optionally, send an image or video clip along with the alert.
Object Recognition
Face Recognition technology is computer vision AI that allows for the identification of specific individuals in real-time video sensors.
Tiny Edge Devices
Xailient’s patented Detectum technology is the worlds’ most efficient AI for Computer Vision. Xailient puts the maximum intelligence into environments with the least size, weight, and power constraints.
Computer Vision (CV) enables camera-equipped systems to understand and act on what they see. Traditionally, CV AI systems require large servers to deliver accurate results. Xailient software, derived from university lab research at Australia’s RMIT, delivers accurate results on tiny devices. Commercial customers use our technology to provide real-time CV on IoT & unconnected low-energy Edge devices.
Our patented Detectum™ technology has been benchmarked by a 3rd party at 509x faster than competitive algorithms for similar tasks. This translates to smaller systems and lower power requirements, reducing the load for soldiers and freeing space for other purposes.
Computer Vision that fits inside of systems reduces or eliminates the need to transmit data. Video data is bandwidth-intensive to transmit. Analyzing imagery at the point of origin (the network’s outer “Edge”) reduces or removes the need to transmit imagery. Systems can be configured to forward only important event data, with or without the associated imagery. Edge CV allows direct integration into alert systems, autonomous systems, or Warfighter situational awareness in real-time. Loitering surveillance systems can operate in an RF silent mode until events require the transmission of an alert, reducing the likelihood of detection by adversaries.
Xailient can address visual spectrum, IR, Hyperspectral, SAR, and other sensor data that produces a 2D ‘image’. Possible applications include Land, Surface/Underwater & Aerial:
Xailient can address visual spectrum, IR, Hyperspectral, SAR and other sensor data that produces a 2D ‘image’.
Computer Vision that fits inside of systems allows imagery analysis at the source, thereby reducing downlink bandwidth use.
Palantir deploys into orbit Xailient’s computer vision AI
For Palantir, Xailient demonstrated an ability to use open-source satellite imagery to detect marine vessels within extremely constrained systems and to do so fast enough to allow for dynamic retasking of the satellite.
Read moreCVOps is a category that describes the enterprise business process Computer Vision in production (Computer Vision OPerationS).
Xailient AI has been validated by 3rd parties as up to 509x faster than conventional Computer Vision AI. We are partnered with Sony Semiconductor and Maxim Integrated, bringing this technology to the private sector. Our software provides faster and more accurate real-time inference, as well as the tools required for creating and maintaining custom AI.
