Anipose: A toolkit for robust markerless 3D pose estimation

TLDR

Quantifying movement is critical for understanding animal behavior, and recent computer‑vision advances enable markerless 2D tracking, though most animals move in 3D. The authors introduce Anipose, an open‑source toolkit for robust markerless 3D pose estimation. Anipose builds on DeepLabCut, providing a 3D calibration module, error‑resolving filters, a triangulation module with temporal and spatial regularization, and a pipeline for large‑scale video processing, along with tutorials and documentation. Anipose was validated on a calibration board and on mice, flies, and humans, and its 3D leg‑kinematics analysis revealed a key role for joint rotation in fly walking motor control.

Abstract

Quantifying movement is critical for understanding animal behavior. Advances in computer vision now enable markerless tracking from 2D video, but most animals move in 3D. Here, we introduce Anipose, an open-source toolkit for robust markerless 3D pose estimation. Anipose is built on the 2D tracking method DeepLabCut, so users can expand their existing experimental setups to obtain accurate 3D tracking. It consists of four components: (1) a 3D calibration module, (2) filters to resolve 2D tracking errors, (3) a triangulation module that integrates temporal and spatial regularization, and (4) a pipeline to structure processing of large numbers of videos. We evaluate Anipose on a calibration board as well as mice, flies, and humans. By analyzing 3D leg kinematics tracked with Anipose, we identify a key role for joint rotation in motor control of fly walking. To help users get started with 3D tracking, we provide tutorials and documentation at http://anipose.org/.