Often security cameras, smart doorbells and locks require battery operation to make them easy to install and to improve security. However, getting a reasonable battery life can present a significant challenge.

Our second generation GAP9 processor revolutionises embedded machine learning in battery powered Smart Home devices. GAP9 wakes up first and checks if the event detected requires further attention. This reduces false positives and resulting network energy and other operational costs.

GAP9 also allows for multi sensor analysis (vision and sound) enabling attractive new features and increasing accuracy.

GAP9 – best in class for battery-powered smart security systems 

Low power rejection of false positives (animals, people walking past)

Multi sensor for anti-spoofing

Low latency and energy inference on image or sound

  • Person detection
  • Face detection / identification
  • Speaker detection / identification
  • Voice driven user interface
  • Abnormal sound detection
  • Sophisticated audio processing for intercom (noise suppression/ beamforming)
  • Ultra fast time to first image – 100ms

State of the art toolchain with a wide range of supported models and frameworks

Small WL-CSP package (3.7×3.7mm)