Noise-Resistant Deep Learning for Object Classification in Three-Dimensional Point Clouds Using a Point Pair Descriptor

Keyphrases

• 3D Point Cloud 100%
• Object Classification 100%
• Deep Learning 100%
• Point Pairs 100%
• Noise-resistant 100%
• Occlusion 66%
• Experimental Validation 33%
• Benchmark Dataset 33%
• Convolutional Neural Network 33%
• Irregular Sampling 33%
• Convolutional Layer 33%
• Class-specific 33%
• Sampling Density 33%
• Point Cloud Data 33%
• Object Retrieval 33%
• Histogram 33%
• Robustness to Noise 33%
• Retrieval Accuracy 33%

Computer Science

• Point Cloud 100%
• Deep Learning Method 100%
• Convolutional Neural Network 50%
• retrieval accuracy 50%
• Convolutional Layer 50%
• Sampling Density 50%
• Information Retrieval 50%