Abstract
Semantic segmentation models help with the extraction of information from images. Currently, Convolutional Neural Networks (CNNs) are the state of the art for performing such tasks but the interpretability in their predictions is low. Previous work had proposed the use of Fuzzy Logic Rule-based systems (FRBS) as an explainable AI classifier of pixels for segmentation of images. In this paper, we extend that approach by using the similarity between image patches as context information for our model. The type-1 FRBS that uses the proposed set of context information features reaches an average Intersection over Union (IoU) value 3.51% higher than the type-1 FRBS using colour information. The difference in average IoU is significant due to the importance of colour in the testing images and the already high IoU value from the type-1 FRBS using colour.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
