Creating a Waterproof IoT Device for Lap Swimmers with Visual Impairments.
Drive the robotic hand to mirror live gestures via NPU-accelerated Qualcomm AI optimized model
Run YOLO26 on the Dragonwing IQ-9075: from 4.9 FPS on the CPU to 153.6 FPS on the NPU, a 31x speedup.
We built a car that autonomously follows lane lines on a track using a camera vision system and PD controller running on a Raspberry Pi 5.
Oracles tell eco-stories & empower action! AI + storytelling gamifies sustainability. Build a greener future with The Nexus!
AI Security on RK3576: YOLOv8 person detection, counting, alerts, snapshots, Flask stream—local, cloud-free.
For the Machine Vision and Intelligence (MVI) assignment, our group developed a vision-based hand gesture recognition system.
Real-time AI system to monitor worker posture and prevent injuries by detecting poor posture and providing corrective feedback.
BeagleBone AI 64 powered autonomous Car that drives and detects blue lane lines and automatically stops at red stop boxes!
Upgrade your robotics with an AI "Super Brain. " ArmPi Ultra fuses LLMs and 3D vision to turn natural language into precise 3D action.
Gesture & AI magic turns play into learning for little explorers—safe, fun and inspiring!
This is a stretching app that uses MoveNet quantized for AMD Ryzen AI to efficiently estimate your pose from the camera.
This session covers an example of multi-core chained processing of a color image using the AMD Ryzen™ AI Phoenix AI engine.
For every bit of damage, you owe a pushup. The code keeps track of the number of times you take damage and keeps your pushup tally for you.
Autonomous RC car with lane detection, PD steering, encoder speed control, and object detection.
This session demonstrates zero padding of an image using Multi dimension capabilities of AIE shim DMA of AMD Ryzen AI Phoenix using AIE-API.
Recent experience troubleshooting Azure Percept
In various contexts such as photography and driver drowsiness, it is import to be able to distinguish when eyes are open or closed.
Capture-the-flag swarm — dual-ESP32 drones (P4 onboard vision, S3 comms), PX4 flight, ROS2 ground station. Open-source.
When you build a tool-using agent that controls a real website, the biggest reliability killer is rarely reasoning quality. It’s context.
Going from demo to production with NVIDIA Deepstream means custom models, C++ pipelines and real world fine-tuning, Find out what we learnt.
A simple app using Raspberry Pi AI Camera and pose estimation to detect when someone is dozing off based on minimal body movement.
HOSTIA: A system to detect falls and other circumstances of daily life that elderly people may suffer at home.
The Owl-tonomous car uses vision and PID to stay on track and glide through curves.
