Enabling Low Power Computer Vision & Audio on Edge Devices
tiny Vision Models
We specialize in developing production quality AI models that are customized to specific applications. These models are trained on use case specific datasets and applications for the best performance with a tiny memory and compute footprint.
This model demonstrates a 10kB, (yes, that right, 10 kilo Bytes) face detector and localizer that is suitable for deployment on a microcontroller.
Custom Audio and Vision Solutions
We work with our customers to develop custom embedded AI solutions that are compact in memory and compute, power efficient, and tailored to your needs. Our expert team of engineers use the latest technology to deliver powerful and responsive solutions. From advanced object detection to noise suppression and echo cancellation, we've got you covered. So why go big when you can go tiny?
Community Driven
At tinyVision.ai, we believe that great things come from working together. That's why we're committed to being community-driven and supporting an open-source philosophy. We believe in sharing our expertise and knowledge with the world, and working collaboratively to build better, more innovative AI solutions that benefit everyone.
Community and Education
At tinyVision.ai, we're all about sparking the curiosity and creativity of the future roboticists and machine learning wizards out there. We sponsor teams in FIRST, RobotSub, and RobotX robotics competitions, offer mentoring and internships, and have introductory FPGA classes. Our goal is to inspire and empower the next generation of engineers and innovators to build our future.