Self-Supervise Reinforcement Learning Method for Vacant Parking Space Detection Based on Task Consistency and Corrupted Rewards

Keyphrases

• Ability to Learn 33%
• Consistency Learning 66%
• Detection Method 33%
• Detection Network 33%
• Dynamic Learning 33%
• False Detection 33%
• Flow-based 33%
• Human-induced Loads 33%
• Learning Methods 66%
• Learning Settings 33%
• Motion Behavior 33%
• Network Target 33%
• Parking Lot 33%
• Performance Improvement 33%
• Reinforcement Learning 33%
• Reinforcement Learning Algorithm 100%
• Reward System 33%
• Self-supervision 100%
• Semantic Consistency 66%
• Space Detectors 66%
• Subspace Detection 33%
• Symmetric Constraint 66%
• Target Network 33%
• Task-oriented 33%
• Training Loss 33%
• Training Samples 33%
• Vacant Parking Space Detection 100%
• Vacant Space 100%

Computer Science

• Detection Method 50%
• Motion Behavior 50%
• Network Target 100%
• Parking Space Detection 100%
• Reinforcement Learning 100%
• Semantic Consistency 100%
• Training Sample 50%