Skip to content

Social LEAP Estimates Animal Poses (SLEAP)

SLEAP Logo

SLEAP is an open-source deep-learning based framework for multi-animal pose tracking (Pereira et al., Nature Methods, 2022). It can be used to track any type or number of animals and includes an advanced labeling/training GUI for active learning and proofreading.

Documentation for New SLEAP Versions

This documentation is for the latest version of SLEAP.
If you are using SLEAP version 1.4.1 or earlier, please visit the legacy documentation.

New in SLEAP 1.5

Check out our What's New page to learn about the latest features including UV-based installation, PyTorch backend, and new standalone libraries!

Features

  • Easy, one-line installation with support for all OSes

  • Purpose-built GUI and human-in-the-loop workflow for rapidly labeling large datasets

  • Single- and multi-animal pose estimation with top-down and bottom-up training strategies

  • Customizable neural network architectures that deliver accurate predictions with very few labels

  • Fast training: 15 to 60 mins on a single GPU for a typical dataset

  • Fast inference: up to 600+ FPS for batch, <10ms latency for realtime

  • Support for remote training/inference workflow (for using SLEAP without GPUs)

  • Flexible developer API for building integrated apps and customization

  • Two independent backends-- sleap-nn and sleap-io for training/inference pipelines & handling SLEAP files respectively

sleap-nn Backend

The SLEAP GUI can be installed and used independently of the sleap-nn backend for labeling. However, for training and inference workflows, it is important that you have sleap-nn installed with the correct PyTorch and CUDA versions according to your machine (ex. CPU or GPU).

Learn more about sleap-nn here.

sleap-io Backend

For working with SLEAP files directly from a CLI, it is best to use sleap-io.

Learn more about sleap-io here.

Get some SLEAP

SLEAP is installed as a Python package. We strongly recommend using uv or Miniconda to install SLEAP in its own environment.

You can find the latest version of SLEAP in the Releases page.

Quick start

Sample with uvx

Note that opening SLEAP w/ uvx will not install SLEAP onto your system, it will only 'invoke' SLEAP.

uvx (any OS):

uvx --from "sleap[nn]" --index-url https://pypi.org/simple --extra-index-url https://download.pytorch.org/whl/cu128 sleap-label

Other CUDA versions

  • For more information on which CUDA version to use for your system, see the PyTorch installation guide. The --extra-index-url in the install command should match the CUDA version you need (e.g., https://download.pytorch.org/whl/cuda118 for CUDA 11.8, https://download.pytorch.org/whl/cuda128 for CUDA 12.8, etc.).
  • On macOS, MPS (Metal Performance Shaders) is automatically enabled for Apple Silicon acceleration.
uvx --from "sleap[nn]" sleap-label

uvx --from "sleap" sleap-label

GUI ONLY

Installing this version of SLEAP will NOT include any training/inference capabilities, as it will not include the sleap-nn backend. This should primarily be used for labeling.

pip (any OS):

pip install sleap

See the docs for full installation instructions.

Learn to SLEAP

References

SLEAP is the successor to the single-animal pose estimation software LEAP (Pereira et al., Nature Methods, 2019).

If you use SLEAP in your research, please cite:

T.D. Pereira, N. Tabris, A. Matsliah, D. M. Turner, J. Li, S. Ravindranath, E. S. Papadoyannis, E. Normand, D. S. Deutsch, Z. Y. Wang, G. C. McKenzie-Smith, C. C. Mitelut, M. D. Castro, J. D’Uva, M. Kislin, D. H. Sanes, S. D. Kocher, S. S-H, A. L. Falkner, J. W. Shaevitz, and M. Murthy. Sleap: A deep learning system for multi-animal pose tracking. Nature Methods, 19(4), 2022

BibTeX

@ARTICLE{Pereira2022sleap,
   title={SLEAP: A deep learning system for multi-animal pose tracking},
   author={Pereira, Talmo D and
      Tabris, Nathaniel and
      Matsliah, Arie and
      Turner, David M and
      Li, Junyu and
      Ravindranath, Shruthi and
      Papadoyannis, Eleni S and
      Normand, Edna and
      Deutsch, David S and
      Wang, Z. Yan and
      McKenzie-Smith, Grace C and
      Mitelut, Catalin C and
      Castro, Marielisa Diez and
      D'Uva, John and
      Kislin, Mikhail and
      Sanes, Dan H and
      Kocher, Sarah D and
      Samuel S-H and
      Falkner, Annegret L and
      Shaevitz, Joshua W and
      Murthy, Mala},
   journal={Nature Methods},
   volume={19},
   number={4},
   year={2022},
   publisher={Nature Publishing Group}
}

Contact

Follow @talmop on X for news and updates!

Technical issue with the software?

  1. Check the Help page.
  2. Ask the community via discussions on Github.
  3. Search the issues on GitHub or open a new one.

General inquiries? Reach out to talmo@salk.edu.

Contributors

  • Talmo Pereira, Salk Institute for Biological Studies
  • Liezl Maree, Salk Institute for Biological Studies
  • Arlo Sheridan, Salk Institute for Biological Studies
  • Arie Matsliah, Princeton Neuroscience Institute, Princeton University
  • Nat Tabris, Princeton Neuroscience Institute, Princeton University
  • David Turner, Research Computing and Princeton Neuroscience Institute, Princeton University
  • Joshua Shaevitz, Physics and Lewis-Sigler Institute, Princeton University
  • Mala Murthy, Princeton Neuroscience Institute, Princeton University

SLEAP was created in the Murthy and Shaevitz labs at the Princeton Neuroscience Institute at Princeton University.

SLEAP is currently being developed and maintained in the Talmo Lab at the Salk Institute for Biological Studies, in collaboration with the Murthy and Shaevitz labs at Princeton University.

This work was made possible through our funding sources, including:

  • NIH BRAIN Initiative R01 NS104899
  • Princeton Innovation Accelerator Fund

License

SLEAP is released under a Clear BSD License and is intended for research/academic use only. For commercial use, please contact: Laurie Tzodikov (Assistant Director, Office of Technology Licensing), Princeton University, 609-258-7256.