VNT-Net: Rotational Invariant Vector Neuron Transformers.

Keyphrases

• Transformer 100%
• Rotation Invariance 100%
• Invariant Vectors 100%
• Transformer Network 100%
• Neural Network 40%
• 3D Point Cloud 40%
• SO(3) 40%
• Attention Layer 40%
• Adaptation 20%
• Challenging Problems 20%
• Machine Learning 20%
• Classification Task 20%
• State-of-the-art Techniques 20%
• Point Cloud 20%
• Superior Performance 20%
• Image Sequence 20%
• Data Augmentation 20%
• Image Patch 20%
• Hyperparameters 20%
• Segmentation Problem 20%
• 3D Point Set 20%
• Transformer Architecture 20%
• Multi-head Attention 20%
• Self-attention Layer 20%
• Cloud Object 20%

Computer Science

• Point Cloud 100%
• Attention (Machine Learning) 100%
• Transformer Network 100%
• Neural Network 66%
• Machine Learning 33%
• Learning System 33%
• Classification Task 33%
• Data Augmentation 33%
• Superior Performance 33%
• Segmentation Task 33%
• Leaning Parameter 33%