LUC3D: Multi-animal 3D pose tracking in the browser

Animal behavior unfolds in three dimensions over time, yet accessible scientific tools for analyzing it in its natural form remain limited. Uncertainty from popular 2D pose methods can be refined with multi-camera approaches to produce more accurate 3D estimates. This paper presents LUC3D (Label Unification and Correspondence in 3D Annotation GUI), a browser-based 3D GUI for reducing the amount of manual labor required for labeling, proofreading, and multi-camera multi-animal tracking while requiring no installation to maximize accessibility. As proof-of-concept we present two datasets of proofread multi-camera 3D triangulated pose data of mouse social behavior, a protocol for generating 3D consistent labels, and a cross-view re-identification tracking algorithm for multi-animal multi-camera pose data. This tool allows for the widespread adoption of 3D pose estimation techniques for studying social behavior while reducing the cost of entry for neuroscience labs across the world.

Authors

Institutions

Publication Details

Journal
bioRxiv (Cold Spring Harbor Laboratory)
Published
2026-11-18
DOI
https://doi.org/10.64898/2026.11.17.767188
Primary Topic
Human Pose and Action Recognition
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

LUC3D: Multi-animal 3D pose tracking in the browser

Isaac Tang, Talmo Pereira, Jinrun Jiang, Liezl Maree et al.
1 links
bioRxiv (Cold Spring Harbor Laboratory)
Human Pose and Action Recognition
preprint

LUC3D: Multi-animal 3D pose tracking in the browser

Isaac Tang, Talmo Pereira, Jinrun Jiang, Liezl Maree, Annegret L. Falkner, Stefan N. Oline, Joshua Park, Bartul Mimica, Kay M. Tye, Mala Murthy, Eric J. Leonardis, Eliza Whitmire, Aaron Zhu, Shayan Afshar, Talisa Wines, Olivia Timmermans
preprint en

Abstract

Animal behavior unfolds in three dimensions over time, yet accessible scientific tools for analyzing it in its natural form remain limited. Uncertainty from popular 2D pose methods can be refined with multi-camera approaches to produce more accurate 3D estimates. This paper presents LUC3D (Label Unification and Correspondence in 3D Annotation GUI), a browser-based 3D GUI for reducing the amount of manual labor required for labeling, proofreading, and multi-camera multi-animal tracking while requiring no installation to maximize accessibility. As proof-of-concept we present two datasets of proofread multi-camera 3D triangulated pose data of mouse social behavior, a protocol for generating 3D consistent labels, and a cross-view re-identification tracking algorithm for multi-animal multi-camera pose data. This tool allows for the widespread adoption of 3D pose estimation techniques for studying social behavior while reducing the cost of entry for neuroscience labs across the world.

bioRxiv (Cold Spring Harbor Laboratory)
Salk Institute for Biological Studies (US), Howard Hughes Medical Institute (US), Princeton University (US)
Howard Hughes Medical Institute, National Institutes of Health, National Institute of Mental Health
Human Pose and Action Recognition
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon