Self-attention-based fully-inception networks for continuous sign language recognition

Keyphrases

• Attention-based 100%
• Self-attention 100%
• Sign Language Recognition 100%
• Inception Network 100%
• Connectionist Temporal Classification 100%
• Feature Extraction 50%
• Validation Set 50%
• Communication Gap 50%
• Phoenix 50%
• Hearing Impairment 50%
• Challenging Tasks 25%
• Feature Learning 25%
• Unsupervised Learning 25%
• Receptive Field 25%
• Normal Hearing 25%
• Overfitting 25%
• Decoding Algorithm 25%
• Pseudo Label 25%
• Feature Extractor 25%
• Sign Language 25%
• Classification Learning 25%
• Elastic Deformation 25%
• Vision-based 25%
• Joint Training 25%
• Cross-entropy 25%
• Limited Datasets 25%
• Cross-entropy Loss 25%
• Self-attention Network 25%
• Sign Word 25%
• Sequential Features 25%
• Frame-level 25%
• Sequential Feature Learning 25%
• Auxiliary Classifier 25%

Computer Science

• Sign Language Recognition 100%
• Self-Attention 100%
• Representation Learning 66%
• Feature Extraction 66%
• Validation Set 66%
• Experimental Result 33%
• Convolutional Neural Network 33%
• Annotation 33%
• Extracted Feature 33%
• Attention (Machine Learning) 33%
• Receptive Field 33%
• Decoding Algorithm 33%
• Classification Learning 33%