Multi- View Fusion for Action Recognition in Child-Robot Interaction.
Niki EfthymiouPetros KoutrasPanagiotis Paraskevas FilntisisGerasimos PotamianosPetros MaragosPublished in: ICIP (2018)
Keyphrases
- multi view
- action recognition
- view invariant
- single view
- multiple views
- human actions
- activity recognition
- depth map
- computer vision
- camera calibration
- semi supervised
- three dimensional
- range images
- body parts
- d objects
- dynamic scenes
- multiple cameras
- human activities
- humanoid robot
- multiple viewpoints
- vision system
- surface reconstruction
- human computer interaction
- viewpoint
- view synthesis
- high resolution
- motion capture
- visual hull