Human Pose and Action Recognition

The paper introduces MARS, a software pipeline for automated analysis of social behaviors in mice, achieving human-level performance in pose estimation and behavior classification.

This paper introduces a video analysis tool that combines supervised pose estimation with unsupervised dimensionality reduction to extract interpretable behavioral features from animal videos, improving downstream analyses.

DeepEthogram uses machine learning to automatically classify animal behaviors from video frames with high accuracy, reducing manual effort and increasing reproducibility.

Anipose provides a toolkit for markerless 3D pose estimation using 2D video, enabling analysis of animal movement and motor control.

B-SOiD is an open-source unsupervised algorithm that identifies spontaneous behaviors from pose data, improving processing speed and generalization across subjects.

This paper develops a monkey pose estimation system using OpenPose, achieving performance comparable to human models and running in real-time.

The paper introduces MacaquePose, a dataset for markerless motion capture of macaques, showing neural network performance comparable to human-level pose estimation.

This paper introduces a real-time, low-latency pose estimation tool with feedback capabilities for behavioral neuroscience experiments.

This paper introduces a deep learning pipeline for tracking and classifying behaviors of interacting primates and mice in complex environments using video data.

OpenMonkeyStudio provides a deep learning-based system for markerless 3D pose estimation in freely moving macaques, enabling automatic tracking and action recognition in unconstrained environments.

This paper evaluates OpenPose's markerless 3D skeleton pose estimates against marker-based motion capture data for pop dance movements, showing average magnitude errors under 700 mm despite self-occlusions.

This study evaluates the accuracy of 3D markerless motion capture using OpenPose with multiple cameras, showing 47% of errors below 20 mm and 80% below 30 mm, but some errors exceed 40 mm due to tracking failures.

FreiPose is a deep learning framework for precise 3D motion capture of animals, showing high accuracy and reliability in tracking rats during electrophysiological and optogenetic experiments.

The paper updates and provides a step-by-step guide for using DeepLabCut, an open-source toolbox for markerless 3D pose estimation across species and behaviors.