Skip to content

readme.md

Latest commit

 

History

History
162 lines (103 loc) · 5.91 KB

readme.md

File metadata and controls

162 lines (103 loc) · 5.91 KB
Raw

FaceStream.AI - Face Recognition for RTSP & MJPEG Cameras

This repository contains the original open-source version of FaceStream.AI and is no longer actively maintained.

FaceStream.AI has evolved significantly since this project was released.

The current generation of FaceStream.AI provides fast, local face recognition for existing RTSP and MJPEG cameras and turns recognized identities into events and automation — without sending camera streams, faces or recognition data to the cloud.

FaceStream.AI 2.x is now available

The new FaceStream.AI is available in multiple editions, including a significantly improved Free edition.

FaceStream.AI Free is the easiest way to get started and does not require an account, e-mail address or licence.

Get the latest FaceStream.AI Free:
https://facestream-ai.com/

For more advanced smart-home, security and building-automation scenarios, FaceStream.AI Pro adds powerful automation and integration capabilities.

What has changed?

Compared with this original repository, the current FaceStream.AI generation has been redesigned and extended with:

  • Modern local face recognition
  • RTSP and MJPEG camera support
  • Multiple camera support
  • Continuous and externally triggered recognition
  • Recognition rules based on who, where and when
  • Detection zones
  • Event and recognition management
  • Smart-home and building automation
  • Webhooks and network integrations
  • MQTT
  • KNX
  • Loxone
  • UDP
  • Syslog
  • E-mail notifications
  • Telegram, Pushover and ntfy
  • Modern web-based configuration and management

Face recognition and image processing run locally on your own hardware. Your camera streams, face images, names and recognition events remain on your system.

Free and Pro editions

Want simple local face recognition for your cameras?

Start with FaceStream.AI Free.

Want to use recognized identities to control your smart home, trigger workflows, integrate security systems or automate a building?

Take a look at FaceStream.AI Pro and the other available editions.

Download, documentation and edition comparison:
https://facestream-ai.com/

About this repository

This repository remains online for existing installations, reference purposes and the history of the FaceStream.AI project.

It represents the original generation of FaceStream.AI and will not receive the features and improvements available in the current product.

For new installations, please use the current FaceStream.AI release:

https://facestream-ai.com/

Face Recognition in Live Video

FaceStream.AI

Screenshots

Features

  • Real-time video streaming with face recognition: Recognize faces in live video and serves a stream with rectangle rendering <your-host>:<5001>/stream
  • Easy configuration with web interface with upload feature of known faces <your-host>:<5000>
  • Eventlog with safed image if a face is recognized, viewable in web interface
  • configurable UDP/HTTP Notification Service for detected faces to notify other services

example homepage

Fast and lightweight for dockerized setups

  • you can adjust the face recognition interval for your needs (default is every 60 frames)
  • uses high performant face detection AI models
  • makes extensive use of threading to use hardware resources efficiently

Installation Guide for FaceStream.AI

This guide provides instructions on how to build and run the Docker image for FaceStream.AI from the GitHub repository.

Prerequisites

Before you begin, make sure you have the following installed:

Cloning the Repository

First, clone the FaceStream.AI repository to your local machine using the following command:

git clone https://github.com/norman-albusberger/FaceStream.AI.git

Building the Docker Image

Navigate to the cloned repository directory:

cd FaceStream.AI

Build the Docker image using the following command. Replace facestream-ai with your preferred image name:

docker build -t facestream-ai .

Running the Docker Image

After the image has been successfully built, you can run it with the following command. Adjust the port mappings as necessary based on the application's requirements:

docker run -p 5000:5000 -p 5001:5001 -v data:/data facestream-ai

Map the ports to your needs. The configuration data, known faces, event log and event images are stored in /data. You could map it to any volume you like.

Verifying the Installation

After running the Docker image, you can verify that the web interface is up and running by accessing it through your browser:

http://localhost:5000

Wenn your input stream is reachable you can access the output stream on:

http://localhost:5001/stream

Replace localhost with your Docker host IP if necessary.

Additional Notes

  • Ensure your Docker daemon is running before executing the build and run commands.
  • Modify the Dockerfile or application code as necessary for custom setups or configurations.

Contributing

Contributions are welcome! Please fork the repository, make your changes, and submit a pull request.

License

FaceStream.AI is licensed under the GNU Affero General Public License v3.0 (AGPLv3).

Key Points of AGPLv3

  • Network Use Is Distribution: Users who interact with the software over a network are afforded the same rights as those who receive binary copies. This means if you run a modified version of the software on a server and users interact with it over the network, you must also share the modified source code under AGPLv3.

  • Share and Share Alike: If you distribute modified versions of the software, you must also make the source code of those versions available under the same license.

  • User Protections: The license