Mask2tasks: Leveraging Segmentation to Enhance Classification Performance in histopathological colorectal Images.
Hieu Le XuanMinh Hoang LeViet V. TruongHuy Phan QuangHoang Vu HuyPublished in: Tiny Papers @ ICLR (2024)
Keyphrases
- pigmented skin lesions
- image classification
- segmentation method
- segmentation algorithm
- test images
- image analysis
- segmentation errors
- fully automatic
- input image
- edge detection
- skin lesion
- image database
- unsupervised segmentation
- image regions
- adaptive thresholding
- image retrieval
- bounding box
- ground truth
- image data
- segmentation accuracy
- color and texture information
- accurate segmentation
- image segmentation
- segmented images
- microscopic images
- multiple objects
- pixel wise
- feature extraction
- image slices
- image features
- piece wise
- grey level
- automatically segmented
- object recognition
- brain tumors
- image segments
- microscopy images
- lung nodules
- feature vectors
- textural features
- text classification
- intensity images
- image processing
- region growing
- grey level co occurrence matrix
- positron emission tomography
- automatic segmentation
- feature space
- training set
- image segmentation algorithms
- image registration
- medical images
- color features
- pixel level
- gabor filters