IR sensor detects objects and sends real-time status to the KiwisIoT dashboard using ESP8266 and Wi-Fi.
My experience using ML/AI with Arduino UNO Q is quite positive (with some glitches). The learning curve is much simpler than Raspberry Pi.
Build a desk-sized robot that detects a human face using on-device ML and steers toward it in real time, all running on the Arduino UNO Q
Squirrels, chipmunks and other creatures are denied while seed is distributed only when birds are recognized by the UNO Q's USB camera!π«πΏοΈ
Deploy YOLOv8 on the Rockchip RK1820 AI accelerator with RKNN3, covering model conversion, runtime setup, and real-time object detection.
AI-powered robot that autonomously tracks and follows people using computer vision and PI control. No programming neededβjust visual blocks.
A beginner's guide to follow along step-by-step.
Build a vision pipeline that detects people, segments each one with a pixel-accurate mask, and blurs them for privacy
Deploy YOLOv11n on Seeed Studio's reComputer RK3576 with the Hailo-8 module β boosting inference from 2.3 FPS to 28.1 FPS.
Long-range AI monitoring with reCamera + Wi-Fi HaLow: real-time video/detection over 1 km, no network needed.
AI Powered Mouse Trap Solution to replace Pure Mechanical Mouse Trap with WebUI for Live Viewing & IoT Notification for capture event alerts
EnerVision transforms buildings into AI smart, energy-efficient spaces through intelligent automation and real-time monitoring.
A System with Arduino UNO Q and Edge Impulse that detects electronic components and announces them by voice for visual impairments people
This project demonstrates that practical object recognition does not always require complex machine learning frameworks.
Build a real-time safety-helmet camera on NE503: train YOLOv8, compile to Hailo HEF, fix an NMS-zero collapse, reach live detection.
Chained MediaPipe Hand Gesture Pipeline on Qualcomm Dragonwing IQ9: Accelerating Multi-Model AI on Qualcomm's Hexagon NPU
Train with Edge Impulse, detect birds on a UNIHIKER K10, and save species counts with two buttons.
Detect Everyday Objects Using Raspberry Pi and Cloud-Based Computer Vision
Train a custom cell model on HUSKYLENS 2, inspect microscope images, and show live detection feedback on UNIHIKER K10 over I2C.
