Randomly Initialized CNN with Densely Connected Stacked Autoencoder for Efficient Fire Detection

Computer Science

• Autoencoder 100%
• color feature 33%
• Computational Complexity 33%
• Computer Vision 33%
• Convergence Speed 66%
• Convolutional Neural Network 100%
• False Alarm Rate 33%
• Fast Convergence 33%
• Feed Forward Neural Networks 33%
• Feedforward Neural Network 33%
• High False Alarm Rate 66%
• Network Architecture 33%
• Slow Convergence 33%

Engineering

• Autoencoder 100%
• Computational Complexity 33%
• Computervision 33%
• Convergence Speed 66%
• False Alarm Rate 100%
• Feedforward 33%
• Fire Detection 100%
• Input Shape 33%
• Limited Performance 33%

Keyphrases

• Very Deep Network 16%