Gesture Recognition

Improve User Interaction Through Gesture Command and Control

Smarter Products Have Smarter Interfaces
Advanced motion sensors like IMUs, radar, and passive IR grid array sensor ICs are finding their way into an ever-growing array of IoT products. This presents exciting new opportunities for utilizing these sensors for smart, gesture-controlled interfaces. Just a few of the devices that can benefit include:
  • Consumer wearables
  • Smartglasses
  • Smarthome appliances and devices
  • Industrial workforce wearables
  • Toys and Gaming Peripherals
SensiML Gesture Recognition Solution
With over 10 years experience in IMU-based gesture recognition, SensiML can provide both turnkey gesture models, custom gestures, and software tools for DIY model development. SensiML's early work and ongoing R&D in gesture recognition have led to the refinement of optimized models that can provide high accuracy in memory footprints as small as 10Kb.

Unlike captive ML solutions and pre-canned models offered by some vendors, SensiML gesture recognition provides unsurpassed hardware flexibility allowing datasets and models to be ported across IMU parts, vendors, and supporting second sourcing strategies for supply assurance. SensiML has been tested and deployed successfully using a variety of sensing technologies and ICs from Coretronic MEMS, Bosch, Infineon, MCube, Rohm, STMicroelectronics, TDK and others.*
Optimized Gesture Recognition Firmware
The demo video you see here was created in little more than a week's effort to showcase a smart-toy concept built around a core gesture recognition library of 6 gestures and negative classification of a corpus of extraneous movements. The model itself used less than 30kB of SRAM and only 7kB of flash, thus making it feasible for very cost-sensitive toy products using 16-bit or even 8-bit microcontrollers.
Platforms and Plans
SensiML has a growing dataset of common navigation and alphanumeric gestures that can be readily applied and tailored to the specific gesture command set needs of a given project. SensiML offers complete Knowledge Pack development services allowing you to focus on your application. Our team can devise a customized project plan using our gesture training data or yours to create a tailored recognition libraries ready to drop into your application.
Platforms and Plans
For project teams with basic machine learning familiarity, desire to learn, or needing to undertake the tasks entirely in-house, SensiML offers its ML Analytics Toolkit suite. A true end-to-end workflow, SensiML Analytics Toolkit supports the complete process from data collection and labeling to Knowledge Pack generation and testing.