End-to-end tutorials

Learn how to build full machine learning pipelines — from raw data to deployed model — using Edge Impulse. These tutorials walk you through every step of the process for your use case or domain: collecting data, designing an impulse, training a model, testing results, and deploying to real hardware.

🖼️ Image tutorials

Explore visual ML applications including image classification and object detection.

• Image classification
• Object detection (bounding boxes)
• Object detection (centroids)

🎧 Audio tutorials

Build models that understand sound — from detecting specific keywords to recognizing environmental noises.

• Keyword spotting
• Sound recognition

📈 Time-series tutorials

Learn how to use motion, sensor, and other temporal data to detect activity, monitor environments, and identify anomalies.

• Motion recognition
• Environmental sensor fusion
• Heart rate and HRV estimation
• Temperature anomaly detection (OPTA PLC)
• OBD automotive data

⚙️ Hardware-specific tutorials

Deploy real applications on supported embedded hardware platforms and development kits.

• Avnet RaSyn Motion Recognition
• Particle Photon 2 – Audio Event Detection
• Particle Photon 2 – Motion Recognition
• Renesas RA8D1 with NVIDIA TAO
• Renesas RZ/V2L Object Detection (X2)
• Sony Spresense Image Classification
• Silicon Labs xG24 Dev Kit – Object Detection
• Syntiant NDP – Keyword Spotting
• Syntiant NDP – Motion Recognition
• TI LaunchXL – Keyword Spotting
• TI LaunchXL – Sound Recognition

🚀 What you’ll learn

By completing these tutorials, you’ll learn how to:

• Collect and label data from devices and sensors
• Design and configure impulses tailored to your data type
• Train and evaluate models using Edge Impulse Studio
• Optimize models for real-time edge deployment
• Deploy and run models on actual embedded hardware

Whether you’re a beginner exploring your first model or an expert integrating with specific hardware, these guides will help you go from idea to deployment, end-to-end.