The Occupancy Management Reference platform uses an infrared sensor and an ultra-low-power processor specifically designed for onboard Artificial Intelligence to provide an easy to install, battery operated occupancy management sensor that works years without maintenance.
Body detection and tracking applications have attracted a lot of interest in video surveillance use cases. But how do you significantly reduce and filter the amount of data processed in a real-time security system? How do you reduce the power consumption of the system? What if one could add a reliable human detection feature that could wake up a surveillance system only when needed. The GAP8, IoT Application Processor makes this possible!
Hearables manufacturers are under great pressure to improve the audio experience with their products and differentiate with environment aware noise reduction (NR) and active noise cancellation (ANC) and new control capabilities such as voice and gesture detection. This is providing challenges with current platforms. GAP processors provide an open, programmable platform that can deliver mW level power consumption in normal operation with peak operation in the 10s of GOPs to run advanced AI-based audio and voice algorithms on a very tiny battery.
From sensors to artificial intelligence, modern hearing devices and personal sound amplification products (PSAP) are packed with technology. They not only help you hear better, but they also help you live a far more fulfilling life. Even if you do not have a hearing problem, you might want an AI wearable that monitors vital signs or selectively filters noise and voices – for instance, allowing you to hear the person across the table in a busy restaurant. The GAP Application Processors make this possible!
GAP processors deliver significant compute ability at an absolute low energy usage point, allowing the integration of sophisticated algorithms like audio capture, keyword control, audio source separation, specific rendering capabilities, etc. into a wireless smart audio device. With GAP processors you can eliminate heat dissipation issues while adding functionality and improving the performance of a smart speaker.
A full open-source Face Identification stack is available for GAP8. The Face Identification algorithm is a combination of a face detector and then a face reidentification engine. This software stack is a perfect demonstration of GAP8’s flexibility in running different algorithms at extremely low power levels.
Head pose estimation can be used in many edge applications where power consumption is a concern. What if whenever a person turns away from the screen, the laptop activates a sleep mode extending its battery life? The ultra-low power application processor GAP8 makes it possible!
GAP processors deliver the computational and power agility necessary to integrate sophisticated signal processing and artificial intelligence into small, battery-operated devices living on the very edge of the network. They can be used for:
Number plate recognition
Vital sign monitoring
High-frequency vibration monitoring
Analog meter (pressure, water, electricity, …) reading
Machine health monitoring
Edge AIoT sensors
Collision avoidance of microdrones
AI-assisted control of smart toys
And other AI and DSP applications where ultra-low power and high peak computing power matter