Deep learning-based automated detection and clustering of potholes using variational autoencoder for efficient road maintenance
dc.contributor.advisor | Chakravorty, Antorweep | |
dc.contributor.author | Grønner, Daniel | |
dc.contributor.author | Eik, Elisabeth | |
dc.date.accessioned | 2023-09-16T15:51:27Z | |
dc.date.available | 2023-09-16T15:51:27Z | |
dc.date.issued | 2023 | |
dc.identifier | no.uis:inspera:129729955:34394921 | |
dc.identifier.uri | https://hdl.handle.net/11250/3089851 | |
dc.description.abstract | In this study we explore the possibility of using unsupervised learning for pothole detection. We will use images of roads from Brazil where both potholes and cracks can be present in the images, as well as clean images with no damage on the road. This will be done using a variational autoencoder (VAE) and clustering. The study will also explore a supervised method, support vector machine (SVM), to compare the performance of supervised model vs. unsupervised model. The goal for this study is to correctly cluster images containing potholes from images that do not contain potholes. | |
dc.description.abstract | ||
dc.language | eng | |
dc.publisher | uis | |
dc.title | Deep learning-based automated detection and clustering of potholes using variational autoencoder for efficient road maintenance | |
dc.type | Master thesis |
Tilhørende fil(er)
Denne innførselen finnes i følgende samling(er)
-
Studentoppgaver (TN-IDE) [866]
Studentoppgaver i informasjonsteknologi, datateknikk / kybernetikk, signalbehandling