arrow
Return

IntegraPose: A unified framework for simultaneous pose estimation and behavior classification

delete2025-10-22
delete0
PRE
AI
F
Farhan Augustine *
S
Sean O’Sullivan
V
Virginia Murray
T
Tatsuya Ogura
W
Weihong Lin
H
Harvey S. Singer
DOI:10.1016/j.neuroscience.2025.10.020delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Unified Pose and Behavior Analysis:IntegraPose delivers pose estimation and behavioral classification within a single inference pass using YOLOv11-Pose model. • High Performance Across Diverse Rodent Datasets:IntegraPose achieves robust performance on four rodent datasets (single- and multi-animal), with rapid inference times (8.3–13.4ms per frame), enabling high-throughput behavioral analysis. • Open-source and User-Friendly: The open-source, GUI-driven toolkit simplifies annotation and analysis, requiring minimal data science expertise, allowing researchers to train custom models for integrated pose-estimation and behavioral classification.

Journal

Journal of Neuroscience cover
Journal of Neuroscience
IF:
4
Papers:
1.4K
Citations:
13.5W

Organization

U
university of maryland baltimore county
Scholars:
3.5K
Papers: 2.6K
Citations: 2
J
Johns Hopkins University School of Medicine
Scholars:
2.1K
Papers: 748
Citations: 2