Edge Impulse

🚀 We’re live at TechEx Events Global in London! February 4–5, Edge Impulse is back at TechEx Global as a Diamond Sponsor of the IoT Tech Expo and the AI & Big Data Expo. Find the team at Stand 129! Stop by to say hello, check out hands-on demos, and explore the latest innovations shaping the future of edge AI. 👉 Learn more: https://lnkd.in/ep8YhfY5 🤝 Attending and want to meet with our AI experts? https://lnkd.in/e4DpTMYY #TechEx #TechConference #TechExGlobal #AIExpo #AIExpoGlobal #IoTExpo #IoTExpoGlobal

Looking forward to a busy frew days at TechEx Events. One of my favorite London event venues.

Our senior devrel engineer Jim Bruges just dropped another entertaining and educational edge AI video about his new project, an Arduino Nesso N1 knock-knock detector. What's that mean? Well, when arriving guests knock on Jim's door rather than ringing the doorbell and he's in a different part of his house, he doesn't always hear it. So he's trained a model that can identify knocking sounds and deployed it to the battery-powered Nesso N1, affixed to his door. When the Nesso N1 detects knocking on the door, its Zigbee integration wirelessly triggers Jim's Home Assistant setup to send him an alert that someone is at the door. Brilliant! From the video description: "Explore how to build a door knock sensor using the Arduino Nesso N1 and Edge Impulse. This video walks through the full process of collecting data and training an embedded AI model to detect door knocks and opening from accelerometer data then deploying it to the Arduino Nesso N1 as an ultra-low-power Home Assistant Zigbee sensor" Watch here: https://lnkd.in/gVyVzVx7 Let us know if you've assembled any smarthome automations with Edge Impulse in the comments below. #homeassistant #Arduino #IoT #edgeAI #zigbee

🚨 Last call webinar alert! 🚨 Be sure to catch our webinar with Atym tomorrow at 9am PST / 12pm EST to learn how to package Edge AI models into portable WebAssembly-based containers that run across CPUs and MCUs — from Linux systems to Zephyr-based microcontrollers. 👉 👉 "From Model to Fleet: Modernizing Embedded AI Deployment with Atym and Edge Impulse" Daniel Kouba and David Tischler will walk you through the process using an #STM32 as well as a #RaspberryPi, and show how easy it is to build this into your existing ML and deployment pipelines. ✅ ( p.s. - We'll make it fun and engaging, promise 😉 ) Signup is quick and easy, here: https://lnkd.in/gKwwiatw #AI #ML #EdgeAI #Linux Jason Shepherd STMicroelectronics Raspberry Pi The Linux Foundation

If you're at #NADA2026 this week, be sure to check out AutoSonixai, powered by Edge Impulse customer GlobalSense Inc.'s edge AI technology. Groundbreaking!

This is shaping up to be a GREAT webinar! 🪁 🐧 Don't miss this one, as we demonstrate how to build an MLOps workflow that can deploy Edge AI models to both MCU's running The Zephyr Project as well as standard Linux devices. Registration is simple, using the link below:

New from our stellar engineering team: OCR for Linux using the Edge Impulse Python SDK. Two-stage approach (text detection + text recognition) runs fully on-device and is suited for automation, document processing, and other vision workloads. Jump in here to get started: https://lnkd.in/d3fVDcwu Nice work Ivan Turasov! #edgeAI #computervision

📬 The January Edition of The Monthly Edge AI Newsletter is here! The new year is off to a great start, and edge AI continues to reshape the technology landscape. This month, we’re sharing what’s ahead as we look toward exciting product innovations in 2026. In this issue, you’ll find: 🚀 Our Campus Ambassador program officially taking flight 🤖 An invitation to join us at Robotic Builders Forum 💼 Two exciting new career opportunities now open 🤝 And more ways to get involved with the Edge Impulse community Stay up to date with the latest innovations, events, and opportunities. Explore what’s new below 👇

New on the Edge Impulse blog: We’ve unlocked hardware acceleration for Android using Qualcomm QNN. Now, models running on #Snapdragon and #Dragonwing devices can leverage the Qualcomm AI Engine Direct SDK (QNN delegate) to route TensorFlow Lite workloads to dedicated AI/DSP engines — delivering significantly lower latency and power compared to CPU-only inference. Benchmarks show ~10× faster inference on quantized models like YOLOv5 and smoother frame times, all with a drop-in integration that keeps your existing Android app architecture intact. This makes on-device ML more real-world ready on mobile and embedded Android platforms, for everything from object detection to complex multi-stage pipelines. Check out the full blog to see how to get started and bring hardware-accelerated ML to your Android projects: https://lnkd.in/dXDd4T28 #edgeAI #tensorflow

Welcome Salman Hafeez to our pre-sales solutions engineering team! Salman is an embedded software engineer with a Ph.D. in controls and a MBA, specializing in real-time, safety-critical systems for robotics and automation. He brings expertise in embedded systems, HMIs, and AI-driven solutions. He is passionate about innovation, collaboration, and driving high-performance embedded systems. Based in the UK, Salman will support our European sales team, contributing technical expertise to consult prospects, design edge AI solutions, and drive success with Edge Impulse excellence. Join us in welcoming Salman to Edge Impulse!