A simple way to run MBARI's Video Annotation and Reference System (VARS), a component of MBARI' s media management system (M3).

• Overview
• Quick Start
• Getting Started
  • Software Prerequisites
  • Next Steps
  • Clean Up
• Applications
• FAQ
  • Registering a Video

MBARI's Video Annotation and Reference System (VARS) consists of a database, a web server, multiple back-end microservices, web applications, and desktop applications.

The m3-quickstart project makes it easy for organizations to spin up everything needed to get started with VARS for video and image annotation.

To start a web server, database, and VARS services, you’ll need Docker installed. Then run:

# Get the source code
git clone https://github.com/mbari-org/m3-quickstart.git  

# Start the web server, database, and VARS services
# The first run takes a while—be patient.
cd m3-quickstart
bin/docker_start.sh

• Docker - Required to launch the VARS microservices
• Python >= 3.7 - Required for supporting scripts. Additional requirements: bin/etc/python/requirements.txt
• ffmpeg - Required for the Python video registration scripts
• Scala - (Optional) Used by some supporting scripts.

• VARS Annotation - The annotation application
• A VARS Friendly-video player. It can be one of the following:
  • Sharktopoda - macOS-only (Recommended)
  • jsharktopoda - cross-platform

1. Open a browser to http://localhost to verify the system is running.
2. Register your first video:
   • Place videos in m3-quickstart/temp/media and browse to them (e.g., http://localhost/media). Note videos can be hosted on other servers and still registered in VARS too.
   • Once you have a video URL, run bin/vars_register_media.sh to register it in the VARS Video Asset Manager
3. Download VARS Annotation and configure it as described in the setup docs.

To shut down the M3/VARS software stack, run:

m3-quickstart/bin/docker_stop.sh

Used to create annotations on video. Runs on Windows, macOS, and Linux. If you nee a build for your OS, request one here.

Requires a VARS-compatible video player (See "Basic Video Annotation").

A web application for searching and retrieving annotations, videos, and images. A link is available on the Welcome to the MBARI Media Management (M3) Server page (http://localhost).

A web application for editing the knowledge base (lexicon and phylogenetic tree of annotation terms). A link is available on the Welcome to the MBARI Media Management (M3) Server page (http://localhost).

VARS Gridview is a bulk editing tool for reviewing and correcting bounding box annotations, such as machine learning generated annotations.

Mondrian is an image annotation application.

M3 stands for "MBARI Media Managment". VARS (the Video Annotation and Reference System) is the largest component of M3.

Before annotating, VARS must know about your video.

1. Make the video web-accessible.
   • If you’ve run docker_start.sh, the included web server can serve videos.
   • Place videos in m3-quickstart/temp/media and browse to them at http://localhost/media.
2. Register the video with VARS.
   • Follow the recommended naming convention.
   • Use the helper script:

cd m3-quickstart
bin/vars_register_media.sh <camera id> <deployment id> <video url>

• camera-id - An identifier for the camera or platform that collected the video.
• deployment id - An identifier for the deployment of the camera.
• video url - The URL to your video. (Do not use localhost; use the machine's hostname instead).

# Example
bin/vars_register_media.sh "Doc Ricketts" "Doc Ricketts 2309" "http://m3.shore.mbari.org/videos/master/2021/11/2309/D2309_20211109T132100.3Z_prores.mov"

Contributions are welcome!

• Found a bug? Open an issue.
• Want to add a feature or fix something? Submit a pull request.
• For larger changes, please start a discussion in the issues first.

When contributing:

1. Fork the repo and create a feature branch (git checkout -b feature/your-feature).
2. Make your changes with clear commit messages.
3. Run tests and ensure everything works.
4. Open a pull request describing your changes.

This project is licensed under the MIT License.