A Novel Trajectory Prediction Method Based on CNN, BiLSTM, and Multi-Head Attention Mechanism

Computer Science

• Convolutional Neural Network 100%
• Bidirectional Long Short-Term Memory Network 100%
• Trajectory Data 100%
• Multi-Head Attention Mechanism 100%
• Experimental Result 33%
• Estimation Accuracy 33%
• Feature Extraction 33%
• local feature 33%
• Shape Information 33%
• Prediction Model 33%
• Temporal Feature 33%
• Key Information 33%
• Comparative Approach 33%
• Term Dependency 33%
• Change Pattern 33%