A Game Theoretic Approach on Semantic Segmentation Along with Transfer Learning

Abstract

Abstract Over the last few years, it has been observed that the segmentation of images is the domain of interest of several researchers. Especially semantic segmentation of images is the region of interest for computer vision and machine learning problems. Many real-world applications like autonomous driving, indoor navigation, and even virtual or augmented reality systems, etc. are developed based on semantic segmentation of images. In this paper, we have analyzed all the existing approaches of segmentation as well as semantic segmentation of images (i.e. semantic segmentation using transfer learning). In the end, we apply a game theoretic approach with transfer learning at the time of pixel classification. At last, we have drawn our conclusion about the state-of-the-art of semantic segmentation using transfer learning and game theoretic approach.KeywordsSemantic segmentationConvolution neural networkTransfer learningGame theoryNash equilibriumCooperative game