Enhancing Dermoscopic Features Classification in Images Using Invariant Dataset Augmentation and Convolutional Neural Networks.
Piotr MilczarskiMichal BeczkowskiNorbert BorowskiPublished in: ICONIP (3) (2021)
Keyphrases
- skin lesion
- feature set
- extracted features
- pigmented skin lesions
- image features
- extracting features
- image classification
- correct classification rate
- feature extraction
- feature vectors
- benchmark datasets
- classification accuracy
- invariant features
- feature representation
- discriminative features
- test images
- image analysis
- convolutional neural networks
- feature space
- input image
- image set
- invariant moments
- image database
- gray level co occurrence matrix
- textural features
- classification method
- image retrieval
- computer aided diagnosis systems
- image collections
- dermoscopy images
- highly discriminative
- computer aided
- handwritten digits
- geometric transformations
- gabor filters
- feature values
- moment invariants
- color features
- image matching
- class labels
- convolutional network
- object recognition
- training set
- benign and malignant
- malignant melanoma
- fisher linear discriminant
- feature selection
- face recognition
- feature points
- feature subset
- skin cancer
- visual features
- cbir systems
- early detection