Dissimilate-and-assimilate strategy for video anomaly detection and localization

Keyphrases

• Abnormal Samples 50%
• Accurate Localization 50%
• Anomalous Events 50%
• Anomalous Objects 50%
• Anomaly Localization 100%
• Area under the Curve 50%
• Challenging Tasks 50%
• Deep Convolutional Autoencoder 50%
• Detection Localization 50%
• Detection Performance 50%
• Dissimilation 50%
• Error Ratio 50%
• Feature Attention Mechanism 50%
• Generalization Ability 50%
• Latent Feature 50%
• Latent Patterns 50%
• Latent Representation 50%
• Learning Representations 50%
• Localization Performance 50%
• Multiple Layers 50%
• Normal Frame 50%
• Normality 50%
• Pedestrian 50%
• Reconstruction Error 50%
• Robust Optimization 50%
• Single Sample 50%
• Spatial Data 50%
• Temporal Data 50%
• Traditional Areas 50%
• Unsupervised Anomaly Detection 50%
• Video Anomaly Detection 100%

Computer Science

• Anomaly Detection 100%
• Attention (Machine Learning) 100%
• Autoencoder 100%
• Detection Performance 100%
• Reconstruction Error 100%
• Robust Optimization 100%
• Sample Distribution 100%
• Spatial Information 100%
• Temporal Data 100%
• Temporal Information 100%
• Unsupervised Anomaly Detection 100%