Tiny AI collaboration targets sound processing
Neuromorphic analog signal processing (NASP) chipmaker POLYN Technology and embedded machine learning development platform provider Edge Impulse say they have partnered to address ultra-low-power on-sensor solutions for wearables, hearables, and Industrial Internet of Things (IIoT). The collaboration, say the companies, aims to leverage POLYN Technology’s NASP platform to enable more customers to develop application-specific chips for one-dimensional signal processing at the sensor level, and help expedite the adoption of ML at the Edge.
“Resource-hungry edge devices present more and more complexity in sound processing and this is an important challenge to address,” says Aleksandr Timofeev, founder and CEO of POLYN. “Additional challenges include power, accuracy, and novel capabilities. The Edge Impulse-POLYN collaboration will help vendors resolve these issues in next-generation products.”
POLYN’s NASP development framework provides fast and easy conversion of trained neural networks from any well-known ML library into neuromorphic analog chips. Neural-Net-To-Chip automation tools provide a fully functional simulation of the resulting math model, converting it into chip production files. These unique tools, says the company, dramatically reduce product time to market, CAPEX, and redesign OPEX.
The POLYN business model provides customer support throughout the entire product development cycle, including neural network selection, training, optimization, and testing on the software simulation. As a result, the size and structure of the neural network is always optimized to the customer’s task, ensuring effective solutions and further investment savings. POLYN’s approach supports fast and cost-effective development of tailored solutions that perform deep learning computations on mass-market devices.
Edge Impulse’s machine learning software is designed to enable enterprises to build smarter edge products, providing powerful automation and low-code capabilities to make it easier to create valuable datasets and develop advanced ML algorithms. The technology empowers developers to bring more ML products to market faster, and helps teams rapidly develop industry-specific solutions in weeks instead of years.
“We’re very excited to work with POLYN to help nurture brand-new solutions using our respective technologies,” says Zach Shelby, CEO and co-founder of Edge Impulse. “The combination of their IP and our platform provides opportunities for users to overcome development challenges while finding valuable new applications for their data.”
The two companies are focusing initially on hearables using POLYN’s combination of voice detection and voice extraction from any ambient environment with the up to 150uW power consumption, enabling novel designs for hearing aids, hearing assistance products, earbuds, and other miniature devices. Additional capabilities, including music detection and processing, are available for that range of products.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
