This repository contains the code for Comparing Deep Learning and Classical Computer Vision for Semantic Segmentation: A comprehensive analysis of cutting-edge techniques and algorithms for precise object segmentation in computer vision tasks. This work was done under the Computer Vision course at IIT Jodhpur.
Kindly find the Youtube video link here. Explaining our approach and results.
Also find here the detailed report of the project.
You need to install the PASCAL-POC dataset into the drive folder, and then seperately downliad the DRIVE dataset(to be able to run the retinal image segmentation) And approptiately change the paths in the code.
Then simply run all the cells in the colab file.
Semantic segmentation is a fundamental task in computer vision that involves assigning a label to each pixel in an image. In recent years, deep learning methods have achieved state-of-the-art performance in semantic segmentation tasks. However, classical computer vision techniques are still widely used and offer several advantages over deep learning methods in certain scenarios. In this project, we aim to conduct a comparative analysis of deep learning and classical computer vision techniques for semantic segmentation