Input-specific Attention Subnetworks for Adversarial Detection.
Emil BijuAnirudh SriramPratyush KumarMitesh M. KhapraPublished in: ACL (Findings) (2022)
Keyphrases
- detection algorithm
- false positives
- detection method
- higher level
- false alarms
- multi agent
- machine learning
- detection accuracy
- automatic detection
- detection rate
- object detection
- domain specific
- real world
- high level
- computer vision
- input data
- digital images
- expert systems
- image sequences
- visual attention
- decision trees
- artificial intelligence
- learning algorithm
- oil spill