BerryNet

BerryNet is a deep learning gateway that turns a Raspberry Pi or similar edge device into a self-contained AI appliance. Developed by DT42, it packages camera input, neural network inference and message publishing into one installable system, so a cheap single-board computer can watch a video feed, run detection models locally and report results to the cloud or a dashboard. It is an open-source project on GitHub with around 1,600 stars.

Overview

Edge AI projects usually require stitching together camera drivers, inference runtimes and IoT messaging by hand. BerryNet provides that pipeline out of the box: it captures images from a Pi camera or USB camera, runs them through TensorFlow, OpenVINO or TensorRT models, and publishes detections over MQTT, the standard protocol for IoT devices. A browser-based dashboard shows the live feed and inference results, and the modular design lets developers swap models and inference engines for different use cases, from people counting to license plate recognition.

Key Features

• Complete edge inference pipeline for Raspberry Pi and similar devices
• Supports TensorFlow, OpenVINO and TensorRT inference engines
• Publishes detection results over MQTT for IoT integration
• Web dashboard for live camera view and inference output
• Modular design: swap models for new detection tasks
• One-command installation on a fresh Raspberry Pi OS image

💳 Pricing

BerryNet is free and open source under the GPL-3.0 license. Costs are limited to your own edge hardware.

Who Is It For

IoT developers, makers and system integrators who need affordable on-device vision AI without sending video to the cloud.