tansor / image-segmentation-by-correlation-clustering Goto Github PK
View Code? Open in Web Editor NEWThis project forked from sheetal158/image-segmentation-by-correlation-clustering
This is the project that we did for the Computer Vision course at Stony Brook University. It is a learning based method where the number of clusters do not have to be specified from the beginning. Image Segmentation by Correlation Clustering gave us extremely good results with Boundary displacement Error as low as 4.85 compared to 10.81 in the work "Higher-Order Correlation Clustering for Image Segmentation"
License: MIT License