A multi-task convolutional neural network with spatial transform for parking space detection

Keyphrases

• Parking 100%
• Loss Function 100%
• Parking Detection 100%
• Lighting Variations 100%
• Spatial Transform 100%
• Object Occlusion 100%
• Perspective Distortion 100%
• Car Size 100%
• Multi-task Cascaded Convolutional Neural Network (MTCNN) 100%
• Outdoor Lighting 100%
• Projection Matrix 50%
• Local Image 50%
• Deep Learning Framework 50%
• Space Images 50%
• Convolutional 50%
• Occlusion Problem 50%
• Occlusion 50%
• Target Space 50%
• System Accuracy 50%
• Status Assessment 50%
• Vision Task 50%
• Image Area 50%
• Image Region 50%
• Handcrafted Features 50%
• Vacant Parking Space Detection 50%
• Camera Geometry 50%
• Camera Projection 50%
• Car Parking 50%
• Non-unified 50%
• Spatial Transformer Network 50%
• Multi-task Loss 50%
• Status Classification 50%
• Error Detection Rate 50%

Computer Science

• Convolutional Neural Network 100%
• Parking Space Detection 100%
• Projection Matrix 50%
• Error Detection 50%
• Learning Framework 50%
• Occlusion Problem 50%
• Detection Rate 50%
• Deep Learning Method 50%
• Transformer Network 50%