3DSMDA-Net: An improved 3DCNN with separable structure and multi-dimensional attention for welding status recognition

Computer Science

• Accuracy 100%
• Attention (Machine Learning) 33%
• Computational Resource 33%
• Control 33%
• Conventional Model 33%
• Convolutional Neural Network 33%
• Discriminative Feature 33%
• Information Sequence 33%
• Interference 33%
• Recognition Accuracy 33%
• Temporal Feature 33%