Research-grade vision tooling

Animal behavior analysis with a cleaner operational surface.

Annolid brings annotation, segmentation, tracking, pose estimation, and behavior analysis into one practical environment for labs that need reproducible results without a brittle pile of disconnected tools.

Install Annolid Open Tutorials Browse Docs

One workspace Annotation, segmentation, tracking, and review in one lab-ready flow.

Human-in-the-loop Correct predictions quickly instead of accepting opaque model output.

Research-ready outputs Keep results structured enough for reruns and downstream analysis.

Frames Annotation Segmentation Tracking Behavior Export

Built for the full analysis loop

The landing page should communicate the same thing as the software: operational clarity, modern computer vision workflows, and tools built for real experiments instead of isolated benchmark demos.

Annotation

Precise labels without workflow drag

Create and refine instance labels, masks, and keypoints with tooling designed for iterative review instead of one-pass labeling.

Segmentation

Practical model workflows

Run segmentation pipelines that support research iteration while keeping outputs and review state understandable.

Tracking

Identity continuity in difficult scenes

Handle overlap, motion, and interaction-heavy videos with workflows that prioritize correction, visibility, and reproducibility.

Choose the path that matches your team

Annolid is useful whether you are curating training data, validating model output, or building a repeatable analysis system for a lab or shared facility.

For labeling teams

Curate high-quality training data

Build clean annotation sets for detection, segmentation, pose, and behavior workflows.

Frame extraction and review
Interactive correction loops
Export-friendly shape semantics

For model operators

Validate pipelines against real experiments

Test inference and post-processing in the same environment where review and comparison actually happen.

Segmentation and tracking evaluation
Video batch workflows
Human-in-the-loop quality control

For labs

Turn ad hoc analysis into a system

Create a repeatable path from acquisition to export so collaborators can reproduce results instead of rebuilding them.

Operational docs and tutorials
Structured outputs for downstream analysis
Research-ready deployment surface

Go straight to the docs surface you need

The landing page stays at annolid.com. Product documentation lives under the docs portal and book paths so users can move directly into the right surface.

Docs portal Open the documentation hub Start from the canonical MkDocs entry point for installation, workflows, deployment, and reference material. Installation Get Annolid running Jump directly to platform-specific setup guidance without going through the full docs index first. Tutorials Follow practical walkthroughs Open curated tutorials for real labeling, segmentation, tracking, and behavior analysis workflows. Reference Find operational details fast Use the reference surface for deployment notes, technical links, and high-signal operational guidance. Deployment Understand the publication model Inspect how landing page, docs, redirects, and external site sync are deployed to the public site. Book Open the Jupyter Book Go directly to the notebook-oriented Annolid book that is published separately under /book/.

Use the landing page as a front door, not a detour.

Start with installation if you want to run Annolid now, or go directly to the docs portal if you already know the workflow you need.

Install Annolid Open Docs Portal Open Book

Cite Annolid

If Annolid supports your work, please cite one or more of the references below.

Yang C, Cleland TA. Annolid: Annotate, Segment, and Track Anything You Need. arXiv:2403.18690 (2024).

@misc{yang2024annolid,
  title={Annolid: Annotate, Segment, and Track Anything You Need},
  author={Chen Yang and Thomas A. Cleland},
  year={2024},
  eprint={2403.18690},
  archivePrefix={arXiv},
  primaryClass={cs.CV}
}

Yang C, Forest J, Einhorn M, Cleland TA. Automated Behavioral Analysis Using Instance Segmentation. arXiv:2312.07723 (2023).

@article{yang2023automated,
  title={Automated Behavioral Analysis Using Instance Segmentation},
  author={Yang, Chen and Forest, Jeremy and Einhorn, Matthew and Cleland, Thomas A},
  journal={arXiv preprint arXiv:2312.07723},
  year={2023}
}

Yang C, Forest J, Einhorn M, Cleland T. Annolid: an instance segmentation-based multiple animal tracking and behavior analysis package (2020).

@misc{yang2020annolid,
  author = {Chen Yang and Jeremy Forest and Matthew Einhorn and Thomas Cleland},
  title = {Annolid: an instance segmentation-based multiple animal tracking and behavior analysis package},
  howpublished = {\url{https://github.com/healthonrails/annolid}},
  year = {2020}
}