# Annolid: Start Here

> Canonical docs now live in the Annolid Docs Portal:
> <https://annolid.com/>

[![Annolid CI](https://github.com/healthonrails/annolid/actions/workflows/CI.yml/badge.svg)](https://github.com/healthonrails/annolid/actions/workflows/CI.yml)
[![Release](https://github.com/healthonrails/annolid/actions/workflows/release.yml/badge.svg)](https://github.com/healthonrails/annolid/actions/workflows/release.yml)
[![Publish to PyPI](https://github.com/healthonrails/annolid/actions/workflows/annolid-publish.yml/badge.svg)](https://github.com/healthonrails/annolid/actions/workflows/annolid-publish.yml)

Annolid is a practical toolkit for annotation, segmentation, tracking, and behavior analysis in research videos.
This book remains available as a compatibility layer and curated walkthrough collection.
For the latest, unified user docs, use the Docs Portal.

## Start Here

1. [Open the Docs Portal](https://annolid.com/)
2. [Install Annolid](https://annolid.com/portal/installation/)
3. [Run the video-first workflow](https://annolid.com/portal/workflows/)
4. [Explore tutorials](https://annolid.com/portal/tutorials/)
5. Use this book for legacy chapter-style walkthroughs when needed.

`extract_frames.md` remains available for workflows where frame sampling is preferred, but it is optional for standard video-first labeling.

## Tutorials and Use Cases

- Place preference analysis
- Behavior classification
- Freezing and motion assays
- DINOv3 keypoint tracking
- Post hoc distance analysis

## CLI and Agent Workflows

- [Expert mode CLI tools](Expert_mode_CLI_tools.md)
- [Prediction review agent workflow](prediction_review_agent.md)

## Learn More

- [High-level overview](high_level_overview.md)
- [Model combinations](combining_models.md)
- [Annolid Zoo](annolid_zoo.md)
- [Video Depth Anything](video_depth_anything.md)

## Community and Citation

- [Get in touch](get_in_touch.md)
- [Contributing](contributing.md)
- [Work based on Annolid](work_based_on_annolid.md)
- [Cite Annolid](cite_annolid.md)
- [License](licence.md)
