An Improved IRNet for Instance Segmentation Based on Image-level Supervision.
Haojiang YaoYa-Li HouRui WangXiaoli HaoZhijiang HouPublished in: CISP-BMEI (2022)
Keyphrases
- segmentation method
- pixel level
- segmentation algorithm
- image segmentation
- grey level
- image analysis
- multiscale
- input image
- edge detection
- test images
- image regions
- image pixels
- image classification
- image data
- segmented images
- energy function
- single image
- image segmentation algorithm
- region growing
- segmentation accuracy
- gray level images
- watershed algorithm
- image content
- berkeley segmentation dataset
- segmentation errors
- image features
- image edges
- gray value
- images of natural scenes
- foreground and background
- joint segmentation
- perceptual grouping
- image segments
- optimal segmentation
- foreground background separation
- low depth of field
- pixel wise
- image representation
- microscope images
- circular hough transform
- gray level
- watershed transformation
- region of interest
- image structure
- high resolution
- multiple objects
- region segmentation
- object segmentation
- image processing
- watershed segmentation
- markov random field
- energy functional
- medical images
- low level
- homogeneous regions
- level set