Improved anomaly detection by training an autoencoder with skip connections on images corrupted with Stain-shaped noise.
Anne-Sophie CollinChristophe De VleeschouwerPublished in: ICPR (2020)
Keyphrases
- anomaly detection
- noise free
- intrusion detection
- detecting anomalies
- gaussian noise
- network traffic
- anomalous behavior
- network intrusion detection
- object recognition
- computer security
- input image
- network anomaly detection
- data mining
- spatial correlation
- one class support vector machines
- network security
- behavior analysis
- detecting anomalous
- detect anomalies
- noise level
- noisy images
- denoising
- machine learning
- data sets
- noise reduction
- input data
- network intrusion
- poisson noise
- knowledge discovery
- neural network