Human-interpretable and deep features for image privacy classification.
Darya BaranouskayaAndrea CavallaroPublished in: CoRR (2023)
Keyphrases
- image features
- extracting features
- extracted features
- image classification
- feature vectors
- feature extraction
- feature values
- feature representation
- textural features
- classification accuracy
- feature set
- static images
- class specific
- feature histograms
- input image
- classification method
- svm classifier
- single image
- spectral features
- test images
- individual features
- visual object recognition
- image content
- multiscale
- image analysis
- classification process
- image data
- low level
- feature space
- classification models
- image representation
- original images
- image matching
- image regions
- confidence scores
- gabor filtering
- keypoints
- color features
- global features
- image retrieval
- invariant features
- text classification
- image pixels
- image set
- feature selection
- support vector machine svm
- single feature
- visual attributes
- similarity measure
- image segmentation
- human observers
- spatial information
- segmentation method
- scale space
- color histogram
- support vector
- feature subset
- classification rules
- classification algorithm
- highly discriminative
- gray level co occurrence matrix
- gray level