The Future is Huge for tinyML
tinyML is a community of researchers and industry engineers focused on bringing Machine Learning capabilities to microcontroller devices.
Arm is a world leader in tinyML technology. With more than 180 billion Arm-based chips shipped, our IP, tools, and 1,100+ software partners are forging the way to help developers build billions of tiny, smart internet of things (IoT) devices. Using foundational intelligence from Arm, these devices improve insight and efficiency across markets, including consumer, medical, automotive, and industrial.
tinyML Applications
Vision
Capture insights from intelligent cameras.
- Object Detection
- Human Pose Estimation
- Depth Estimation
Voice
Enable key word detection and automated speech recognition locally on the device—with no cloud required.
- Key Word Spotting (KWS)
- Automatic Speech Recognition (ASR)
- Beamforming
- Noise Suppression
- Machine Translation
Vibration
Leverage vibration to analyse signals, monitor health, predict maintenance and detect anomalies.
- Human Activity Recognition
- Industrial Anomaly Detection
- Sensor Fusion
- Motor Control
- Predictive Failure
Related Products and Services
Cortex-M55 Processor
Arm’s most AI-capable Cortex-M processor delivers enhanced, power-efficient DSP and ML performance.
Ethos-U55 NPU
A new class of machine learning (ML) processor, called a microNPU, specifically designed to accelerate ML inference in area-constrained embedded and IoT devices. The Ethos-U55 combined with the AI-capable Cortex-M55 processor provides a 480x uplift in ML performance over existing Cortex-M based systems.
Ethos-U65 NPU
The Ethos-U65 microNPU ushers in a new wave of AI on the edge and endpoint by supporting ML on Cortex-A based systems and increasing performance on Cortex-M based systems.
Software and Tools
The combination of Arm Keil MDK and Mbed provides a platform for building and deploying tinyML devices.
tinyML Resources
Videos
- Introduction to tinyML
- TinyML with TensorFlow and Mbed OS
- Machine learning made possible for embedded developers
- A low code approach to secure and productive AIoT application
Workshops
- tinyML development with TensorFlow Lite for Microcontrollers
- Simplified deep learning using OpenMV Cam, Edge Impulse and Arduino
Webinars
- Automatically Build tinyML Solutions on Embedded Devices
- Deploy ML Models Faster on Arm Using Open Source \ Apache TVM with OctoML
Blog Posts
Software Developer Resources
Development is simple and accessible with our combination of easy-to-use software, hardware, and tools from our AI partners. Explore resources to get started.