Automated pose estimation in 3D point clouds applying annealing particle filters and inverse kinematics on a GPU

Keyphrases

• 3D Point Cloud 100%
• 3D Sensor 50%
• Annealing 100%
• Body Model 50%
• Body Parts 50%
• Camera System 50%
• Depth Image 50%
• Detection Method 50%
• Full-body Tracking 50%
• GPU-based 50%
• High Demand 50%
• Image-based Detection 50%
• Interaction Paradigms 50%
• Inverse Kinematics 100%
• Kinematic System 50%
• Marker Tracking 50%
• Motion Features 50%
• Natural Interaction 50%
• Particle Filter 100%
• Point Cloud 50%
• Pose Estimation 100%
• Processing Power 50%
• Real-time Performance 50%
• Recent Advances 50%
• Systems-based 50%
• Tracking System 50%
• Virtual Environment 50%

Computer Science

• Detection Method 50%
• Graphics Processing Unit 100%
• Human-Computer Interaction 50%
• Inverse Kinematics 100%
• Motion Feature 50%
• Natural Interaction 50%
• Particle Filter 100%
• Point Cloud 100%
• Pose Estimation 100%
• Processing Power 50%
• Real Time Performance 50%
• Speed-up 50%
• Tracking System 50%
• Virtual Reality 50%

Engineering

• Based Detection Method 50%
• Body Model 50%
• Body Part 50%
• Camera System 50%
• Graphics Processing Unit 100%
• Inverse Kinematics 100%
• Particle Filter 100%
• Point Cloud 100%
• Pose Estimation 100%
• Processing Power 50%
• Test Subject 50%
• Virtual Reality 50%