PointFace: Point Cloud Encoder-Based Feature Embedding for 3-D Face Recognition

Keyphrases

• 2D CNN 50%
• 2D Face Recognition 25%
• 2D Map 50%
• 3D Data 25%
• 3D Face 25%
• 3D Face Recognition 100%
• Bosphorus 25%
• CNN-based 25%
• Deep Learning Methods 25%
• Encoder 100%
• Extracting Features 25%
• Face Recognition Accuracy 25%
• Face Recognition Algorithm 25%
• Face Representation 25%
• Feature Embedding 100%
• Feature Extraction 25%
• Feature Similarity 25%
• Large-scale Training 25%
• Lightweight Framework 25%
• Negative Face 25%
• Pair Selection 25%
• Point Cloud 100%
• Positive Face 25%
• Selection Strategy 25%
• Set Data 25%
• Shared Encoder 25%
• Similarity Loss 25%
• Training Data 25%
• Two-weight 25%

Computer Science

• 2D Face Recognition 33%
• 3d Face Recognition 100%
• Convolutional Neural Network 66%
• Deep Learning 33%
• Face Recognition 100%
• face recognition algorithm 33%
• Face Representation 33%
• Feature Extraction 33%
• Point Cloud 100%
• Recognition Performance 33%
• Selection Strategy 33%
• Training Data 33%