OpenPose: Real-time multi-person keypoint detection library for body, face, hands, and foot estimation

Code repo for realtime multi-person pose estimation in CVPR'17 (Oral)

Keras version of Realtime Multi-Person Pose Estimation project

Training repository for OpenPose

Code for the Pose Residual Network introduced in 'MultiPoseNet: Fast Multi-Person Pose Estimation using Pose Residual Network' paper https://arxiv.org/abs/1807.04067

Pytorch version of Realtime Multi-Person Pose Estimation project

[CVPR 2025] Official implementation of three papers "Certified Human Trajectory Prediction", "Vehicle trajectory prediction works, but not everywhere", and "Are socially-aware trajectory prediction models really socially-aware?".

Code for the paper "Generative Adversarial Graph Convolutional Networks for Human Action Synthesis", WACV 2022

AI Watch A1: Multi-person 3D skeleton detection using Intel RealSense and OpenPose with Kafka support.

Realtime Multi-Person Pose Estimation data server. Used as a training and validation data provider in training process.

[ICCV2023] Chaotic World: A Large and Challenging Benchmark for Human Behavior Understanding in Chaotic Events

Apache Spark utilities to analyze mobile network deployment and user behavior data.

This is an long-term Aritificial Intelligence research project thai I'm working on. The project is focused on developing a comprehensive AI system that can understand and analyze various aspects of human behaviour.

The notes of various MOOCs that I took and continue to take over the years.

Dataset and Codebase for "GREEN PATH: an expert system for space planning and design by the generation of human trajectories".

An Agentic AI Platform for Synthesizing and Observing Human Persona Behaviors !

Minimalist virtual platform for joint exploration for the study of perceptual skills in contexts of social interaction. Used by Researchers in Psychology in order to study social interaction behaviors in humans.

A web-based simulation of human behavior using distinct agent types (Leaders, Influencers, Rebels, Observers) with evolving emotions. Explore social dynamics by adjusting parameters like influence and rebellion. Features AI-powered analysis via Google Gemini API to interpret collective trends and emotional shifts.