Bi-RDNet: Performance Enhancement for Remote Sensing Scene Classification with Rotational Duplicate Layers

Keyphrases

• Performance Enhancement 100%
• Remote Sensing Image Classification 100%
• Expert Group 50%
• Power Constraint 50%
• Layer Parameters 50%
• Network Layer 50%
• Network Accuracy 50%
• Field Programmable Gate Arrays 50%
• Design Decisions 50%
• Limited Availability 50%
• Convolutional Layer 50%
• Effective Network 50%
• Computer Vision Tasks 50%
• Resource-constrained Devices 50%
• Duplication 50%
• Fast Evaluation 50%
• Network Power 50%
• Partial Decompression 50%
• Evaluation Time 50%
• Network Memory 50%
• Compact Network 50%
• Model Size 50%
• Accuracy Constraints 50%
• Memory Saving 50%
• Forward Pass 50%
• Regular Convolution 50%
• Memory Constraints 50%

Computer Science

• Speed-up 100%
• Convolution Layer 100%
• Computer Vision Task 100%
• Performance Enhancement 100%
• Network Layer 100%
• Compressed Data 100%
• Power Constraint 100%
• Field Programmable Gate Arrays 100%