AI - Edge

Computer Vision with Arduino

train on the cloud

SDK

https://docs.edgeimpulse.com/tutorials/tools/sdks/studio/python/upload-download-data

Audio

Audio classification https://docs.edgeimpulse.com/tutorials/end-to-end/sound-recognition

Images

https://docs.edgeimpulse.com/tutorials/end-to-end/object-detection-centroids

FOMO (Faster Objects, More Objects) is a brand-new approach to run object detection models on constrained devices.

FOMO is a ground-breaking algorithm that brings real-time object detection, tracking and counting to microcontrollers for the first time. FOMO is 30 times faster than MobileNet SSD and can run under 200K of RAM.

https://studio.edgeimpulse.com/public/783690/latest

Data adquisition:

  • Labeling queue
  • AI labeling

Impulse desing

  1. Create imoulse input size!! image data. -> processing block -> learning block -> output features

An impulse takes raw data, uses signal processing to extract features, and then uses a learning block to classify new data.

Image

  1. Deploymnet

Options … Set the target device,

Computer Vision

  1. Louis Moreau – Head of Developer Relations at Edge Impulse
  2. Jim Bruges – Senior Developer Relations Engineer at Edge Impulse
  3. Marc Pous – Developer Relations Staff Engineer at Edge Impulse

Hands-on workshop to explore Computer Vision on the edge using Edge Impulse and Arduino UNO Q.

Learn how to collect data and train an object detection model to filter events in real time, then cascade to a Vision Language Model.

You’ll build an end-to-end cascaded AI application in the edge, balancing power efficiency and accuracy on real hardware.

https://app.arduino.cc/courses

Not adapt rules, collect more data.

Docker: training with TensorBoard Model created with Tensorflow Lite

Traditional object detection models are poorly suited to MCUs. (Microcontrollers)

Object detection using centroids
FOMO: Faster objects, more objetcs

MobileNet V2

https://github.com/edgeimpulse/example-arduino-app-lab-object-detection-using-flask.git cd example-arduino-app-lab-object-detection-using-flask/ laptop and arduino on same network.

Client

https://github.com/edgeimpulse/mobile-client