raw signal detection
Physical layer-based AI module for hyper resolution, enhanced SNR and target-specific detection
Target classification for robust and detailed multi-class identification.
Superior object tracking
State of the art prediction models and combined multi-feature fusion for object tracking.
The next generation radar
Uncover the uncompromising potential of AI
Deep Learning technologies can not only improve performance of existing radar tasks, but also enable entirely new capabilities and operational concepts.
DATA DRIVEN RADAR
REDUCED POWER CONSUMPTION
Learning models from raw data can optimize the physical-layer and signal-processing design; Thus, making it possible to reach levels of optimization that were previously not reachable.
Harnessing AI parallelization aptitude allows inference optimization on the edge to perform signal processing and millisecond delay achieve detection and classification from raw data.
When compared to traditional methodologies, AI can show several orders of magnitude of performance improvement. These gains translate to reduced transmitter power and computational requirements.
Defense and beyond
Extended range and low RCS detection with hyper resolution localization.
Noise and external signals removal.
Detailed and robust fine-grained target classification.
Near clutter and adjacent targets differentiation.
Weight load reduction for light mobile platforms as drones and autonomous vehicles.
Radiation minimization for nearby human safety in healthcare.
Adaptive configuration to surrounding interruptions.
Fine pattern recognition for human-machine interface.
Radar classification challenge winner
MAFAT Radar Challenge: Solution by Axon Pulse
Radar target classification — is it a human or an animal?
About Axon Pulse
Axon Pulse revolutionizes the radar industry using AI and Deep Learning techniques.
We present the next-generation radar signal processing using Deep Neural Networks, for increased range and sensitivity, detection super-resolution, and tracking capabilities.
Our products are integrated into major systems in the defense, healthcare, automotive, and drones sectors, through partnerships with leading multinational companies.
Traditional techniques do not use any a-priori knowledge about the radar signals structure, thus imposing a hard limit on detection ranges and resolutions. Superior SNR and resolution used to be achievable solely by hardware upgrades.
We uncover the full and uncompromising potential of AI to solve today’s most critical challenges in radar signal analysis, including end-to-end I/Q-based object detection and classification, object re-identification and appearance-based tracking, and sensor fusion. This allows us to reach previously impossible levels of performance, leading tomorrow’s AI-powered radar technology.