“What's This?” - Learning to Segment Unknown Objects from Manipulation Sequences

Keyphrases

• Unknown Object 100%
• Manipulator 100%
• Training Data 50%
• Kinematics 50%
• Agnostic 50%
• Labeled Data 50%
• Sensor Data 50%
• 3D Object Models 50%
• Semantic Knowledge 50%
• Motion Semantics 50%
• Foreground Segmentation 50%
• Visual Registration 50%
• Hand-eye Calibration 50%
• Semantic Segmentation Network 50%
• Motion Cue 50%
• RGB Frames 50%
• Additional Sensor 50%
• Object Segmentation 50%
• Trainable 50%
• Static Scene 50%
• Robotic Manipulator 50%
• Single End 50%

Computer Science

• Manipulator 100%
• Image Segmentation 33%
• Detailed Insight 33%
• Training Data 33%
• Semantic Knowledge 33%