Team
Abdeali Harianawala, Bhavesh Waghela, Keyur Finaviya, Siri Karnala
Categories
Artificial Intelligence
Overview
The Kevares - Sidewalk Detection and Segmentation Project by Fusion Force is driven by the imperative need to elevate sidewalk analysis and navigation capabilities utilizing advanced deep learning algorithms. Our primary motivation stems from the desire to enhance urban infrastructure maintenance and safety through the development of a sophisticated sidewalk segmentation model for the Kevares Team.
Throughout this project, our team has achieved significant milestones, including the successful annotation of over 1500 sidewalk images and the exploration of various deep learning architectures such as YOLO, Unet, and SAM. These achievements have paved the way for enhancing the efficiency and accuracy of sidewalk analysis, ultimately contributing to the seamless integration of autonomous sidewalk inspection technology for Kevares.
Semantic Segmentation of Sidewalk - Kevares
The video demonstrates the work done for the Capstone Project with Kevares Autonomous Solutions. A Deep Learning model, the Yolo v8 segmentation model performs Sidewalk Segmentation from the captures done by a Robot Mark 3 running on the sidewalks.
