The AIS-C100A Mantis chip is a 96×96 AI-in-Sensor TinyML chip especially fitted for classification tasks. It includes a programmable always on wake-up CNN engine which can be used to turn on external microcontrollers when an object of interest is detected, as well as powerful post wake-up ML capability. All supporting circuitry is provided including power management, timing, AI, and communications as well as an up to 40mA LED driver offering both linear and PWM control on a single monolithic die with a minimum of external components required to save space. Communication is through an SPI port. The exposure time can be internally or externally controlled to reach best possible contrast for AI calculations. Mantis can also process information loaded through the SPI port in addition to the on-chip imager. The AIS-C100A is packaged in a 4.6×5.5 mm OLGA package.
Mantis is a unique machine learning solution that includes all the components required to implement a low cost camera solution.
96×96 Pixel AI-in-Imager
Low power sensor pixel imager with on-chip integrated AI processing
< 80μW programmable always on monitoring when looking for an object
True “always on” where the pixels are part of the neural network
4.6×5.5mm OLGA Package
SPI Interface
– load weights & configuration
– upload or download an image
– read system status
On board power management with on chip wakeup sequencer for external system wakeup
Programmable GPIOs
LED Driver: Current and PWM programmable up to 40mA
Single 3.3V supply or optional external 1.8V supply input
Digital control I/O lines configurable for 1.8V or 3.3V operation
On Chip Oscillator
Built in programmable AI functionality for both always on and post wakeup
Multiple Starter Models:
– face detect
– people counting
– male vs. female
– classification
– occupancy detection
– animal vs. person
– age
People Counting · Face Detection · IR Beam Replacement · Behavior Monitoring · Activity Monitoring ·
Wellness Monitoring · Smart Cities · Eye Tracking · Key Word Spotting · Pupil Dilation · Noise Monitoring · Sound Identification · Vibration Monitoring · Object Detection
Automated page speed optimizations for fast site performance