# DeepLabCut

> software toolbox for markerless pose estimation

**Wikidata**: [Q105177558](https://www.wikidata.org/wiki/Q105177558)  
**Source**: https://4ort.xyz/entity/deeplabcut

## Summary
DeepLabCut is a software toolbox designed for markerless pose estimation of animals, utilizing deep learning techniques to track movement without physical markers. Initiated in April 2018, this open-source project is widely used in computer vision and animal testing research. It is funded by the Chan Zuckerberg Initiative and operates under a GNU LGPL v3.0 license.

## Key Facts
- **Entity Type:** Software toolbox, instance of open-source software.
- **Primary Function:** Markerless pose estimation.
- **Core Technology:** Deep learning (Q21447895).
- **Applications:** Computer vision, animal testing.
- **Inception Date:** April 2018.
- **License:** GNU Lesser General Public License, version 3.0.
- **Funding:** Chan Zuckerberg Initiative (Essential Open Source Software for Science program grant).
- **Copyright Status:** Copyrighted.
- **Official Website:** http://www.mousemotorlab.org/deeplabcut
- **Source Code Repository:** https://github.com/DeepLabCut/DeepLabCut
- **Twitter Handle:** @DeepLabCut (ID: 1096406544676282368).
- **Social Media Metrics:** 9,006 followers (as of March 14, 2023).
- **Latest Stable Version:** 2.3.11 (Released February 18, 2025).

## FAQs
### Q: What specific technologies does DeepLabCut use?
DeepLabCut utilizes deep learning techniques to perform markerless pose estimation. It is categorized under open-source software and is applied primarily in fields requiring computer vision and animal testing.

### Q: Who funds the development of DeepLabCut?
The project receives funding from the Chan Zuckerberg Initiative. Specifically, it is supported through the "Essential Open Source Software for Science program grant."

### Q: What are the licensing terms for DeepLabCut?
The software is copyrighted but released as open-source under the GNU Lesser General Public License, version 3.0. This allows users libre access to the original source code, fitting the definition of open-source software which permits free use and redistribution.

### Q: When was DeepLabCut created and last updated?
The project was inceptioned in April 2018. Its most recent stable release as of the provided data is version 2.3.11, which was published on February 18, 2025.

## Why It Matters
DeepLabCut serves as a critical tool in the scientific community, bridging the gap between advanced artificial intelligence and biological research. By enabling markerless pose estimation, it solves the complex problem of tracking animal behavior and motion without the need for intrusive physical markers, which can alter natural behavior.

Its classification as open-source software amplifies its importance; it embodies the principles of "open knowledge" by offering free access to its source code, thereby lowering barriers to entry for researchers and promoting transparency. The software has been recognized as a prominent example of AI and machine learning applications in science, cited alongside other significant open-source projects. Funding from major philanthropic organizations like the Chan Zuckerberg Initiative highlights its perceived value in advancing essential scientific software infrastructure.

## Notable For
- **Markerless Technology:** Distinguishing itself as a specialized toolbox for "markerless" pose estimation, a significant advancement over traditional motion capture methods.
- **Open Source AI:** Being a prominent example of open-source software specifically within the AI and Machine Learning domain, cited alongside projects like Ray and EleutherAI.
- **Scientific Community Impact:** Receiving specific grant funding from the Chan Zuckerberg Initiative's "Essential Open Source Software for Science program," signaling its utility in high-level research.
- **Active Development Cycle:** Maintaining a rapid and consistent release schedule from its inception in 2018 through 2025, with frequent updates and patches (e.g., moving from version 1.0 to 2.3.11 in roughly seven years).

## Body
### ### Definition and Classification
DeepLabCut is a software toolbox explicitly designed for markerless pose estimation. It is formally classified as an instance of **open-source software** (OSS). In the context of knowledge taxonomies, OSS is defined as software that anyone is free to use and redistribute in its current state, granted through a permissive license that provides "libre access" to the original source code. DeepLabCut fits this definition by operating under the **GNU Lesser General Public License, version 3.0**, though the source material notes that it retains a copyrighted status.

The software is categorized as a component of "open knowledge" and is considered the opposite of proprietary software. It is listed as a prominent example of AI and Machine Learning software, alongside tools like Ray, EleutherAI, Feast, and the NSFW Filter.

### ### Technical Functionality and Use
The primary utility of DeepLabCut is to perform **markerless pose estimation**. This process relies on **deep learning** (identified by the Wikidata ID Q21447895) to analyze video data. The software is utilized primarily in two domains:
1.  **Computer Vision:** Providing the technical backbone for tracking and estimating poses.
2.  **Animal Testing:** Serving as a tool for monitoring and analyzing animal behavior without the need for physical markers.

### ### History and Development Timeline
The project was officially inceptioned in **April 2018**. Its development has been characterized by a high frequency of updates and version releases managed via its GitHub repository.

**Release History Highlights:**
The development timeline shows a progression from initial alpha/beta stages to stable long-term support. The first recorded version in the provided data is **1.0** (released August 20, 2018), followed quickly by patches 1.01 and 1.02. The software moved into the **2.x** generation in January 2019 with version 2.0.3.

Subsequent years saw numerous minor and major updates:
-   **2019-2020:** Versions 2.0 through 2.1.7.1 were released, refining the core toolbox.
-   **2021:** The project released versions 2.1.10 through 2.2.0.3, maintaining monthly or bimonthly update cadences.
-   **2022:** Updates continued with versions 2.2.0.4 through 2.2.3.
-   **2023:** The project moved to version **2.3** in late December 2022, with subsequent stable releases (2.3.2 to 2.3.8) throughout 2023.
-   **2024-2025:** Recent development focused on stability, with releases 2.3.9 and 2.3.10 in 2024, culminating in the latest stable release, **2.3.11**, on **February 18, 2025**.

### ### Financial Support and Community
DeepLabCut is financially supported by the **Chan Zuckerberg Initiative**. The funding is specifically attributed to the "Essential Open Source Software for Science program grant," indicating its role as a critical piece of scientific infrastructure.

The project maintains an active online presence:
-   **Website:** Hosted at `http://www.mousemotorlab.org/deeplabcut`.
-   **Social Media:** It maintains a Twitter account (@DeepLabCut) with a following of **9,006 users** as of March 14, 2023. The account was created on February 15, 2019.
-   **Repository:** The source code is actively maintained on GitHub under the user `DeepLabCut`.

## References

1. [Release 1.0. 2018](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v1.0)
2. [Release 1.1. 2018](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v1.1)
3. [Release 1.01. 2018](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v1.01)
4. [Release 1.02. 2018](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v1.02)
5. [Release 1.11. 2018](https://github.com/DeepLabCut/DeepLabCut/releases/tag/1.11)
6. [Release 2.0.3. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.3)
7. [Release 2.0.4. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.4)
8. [Release 2.0.4.1. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.4.1)
9. [Release 2.0.5. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.5)
10. [Release 2.0.6. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.6)
11. [Release 2.0.6.2. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.6.2)
12. [Release 2.0.7. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.7)
13. [Release 2.0.9. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.0.9)
14. [Release 2.1. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1)
15. [Release 2.1.1. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.1)
16. [Release 2.1.4. 2019](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.4)
17. [Release 2.1.5. 2020](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.5)
18. [Release 2.1.5.2. 2020](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.5.2)
19. [Release 2.1.6. 2020](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.6)
20. [Release 2.1.6.4. 2020](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.6.4)
21. [Release 2.1.7.1. 2020](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.7.1)
22. [Release 2.1.10. 2021](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.10)
23. [Release 2.1.10.2. 2021](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.10.2)
24. [Release 2.1.10.4. 2021](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.1.10.4)
25. [Release 2.2.0.3. 2021](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.0.3)
26. [Release 2.2.0.4. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.0.4)
27. [Release 2.2.0.5. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.0.5)
28. [Release 2.2.0.6. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.0.6)
29. [Release 2.2.1. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.1)
30. [Release 2.2.1.1. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.1.1)
31. [Release 2.2.2. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.2)
32. [Release 2.2.3. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.2.3)
33. [Release 2.3. 2022](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3)
34. [Release 2.3.2. 2023](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.2)
35. [Release 2.3.4. 2023](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.4)
36. [Release 2.3.5. 2023](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.5)
37. [Release 2.3.6. 2023](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.6)
38. [Release 2.3.7. 2023](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.7)
39. [Release 2.3.8. 2023](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.8)
40. [Release 2.3.9. 2024](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.9)
41. [Release 2.3.10. 2024](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.10)
42. [Release 2.3.11. 2025](https://github.com/DeepLabCut/DeepLabCut/releases/tag/v2.3.11)
43. [Source](https://chanzuckerberg.com/eoss/proposals/)