Action recognition by multiple features and hyper-sphere multi-class SVM

Keyphrases

• Multi-feature 100%
• One-class Support Vector Machine (OCSVM) 100%
• Action Recognition 100%
• Hypersphere 100%
• Global Features 66%
• Single Feature 66%
• Statistical Model 33%
• Publicly Available Dataset 33%
• Recognition-based 33%
• Local Features 33%
• Global Information 33%
• Human Action 33%
• Cuboid 33%
• Human Action Recognition 33%
• Individual Size 33%
• Local Space-time Features 33%
• Video Representation 33%
• Spatio-temporal Volume 33%
• Feature-based Representation 33%
• Higher-order Statistics 33%

Computer Science

• Support Vector Machine 100%
• Action Recognition 100%
• Global Feature 66%
• Single Feature 66%
• Scale-Invariant Feature Transform 33%
• local feature 33%
• Points of Interest 33%
• Global Information 33%
• Statistical Model 33%
• Video Representation 33%