It offers an array of distinct features, such as models with an average size of less than 1 KB and the ability to run on 8-bit microcontrollers.
Key Features of Neuton TinyML :
- No-Code Automation: Neuton offers a no-code approach, making it user-friendly and efficient for developers without extensive coding expertise.
- Tiny Model Sizes: The platform excels at creating incredibly small machine learning models, often less than 1 KB in size, ensuring efficient use of resources on edge devices.
- Selective Connections: Neuton TinyML adopts a selective approach to establishing necessary neuron connections, reducing randomness in model creation and enhancing efficiency.
- Silicon Agnostic: It allows models to be deployed on a wide range of microcontrollers, small computing devices, and sensors, making it adaptable to diverse hardware.
- Zero Compression: Neuton TinyML maintains the original characteristics of models without relying on reduction techniques such as quantization, pruning, clustering, or distillation.
- Free Developer Plan: A free plan is available for developers, offering access to the platform at no cost.
- Extensive Benchmarking: It has demonstrated superior performance in benchmark tests when compared to traditional models, making it a compelling choice for edge device AI development.
- Smartwatch Gesture Interaction: Recognizes gestures for smartwatches and wearables.
- Package Tracking: Identifies events during the package delivery process.
- Oral Hygiene Tracking: Monitors teeth-brushing effectiveness.
- Remote Control: Controls media systems through gestures.
- Daily Activity Recognition: Efficiently recognizes complex human activities.
Neuton TinyML is a groundbreaking solution for edge devices, offering the ability to embed extremely compact and accurate models without coding, compression, or additional complexity.
With applications spanning from smartwatches to package tracking, its impressive performance and ease of use make it a valuable asset for developers seeking to bring intelligence to the edge.
The partnership with STMicroelectronics underscores the growing recognition and potential of Neuton’s unique approach in the field of ultra-tiny neural networks.