Joint Facade Registration and Segmentation for Urban Localization
This work addresses urban localization challenges for applications like autonomous navigation or mapping, but it is incremental as it builds on existing facade detection and segmentation methods.
The paper tackles the joint problem of facade registration and semantic segmentation in urban images by proposing a Bayesian model within an Expectation-Maximization framework, resulting in improved accuracy and robustness to clutter and illumination changes across various databases.
This paper presents an efficient approach for solving jointly facade registration and semantic segmentation. Progress in facade detection and recognition enable good initialization for the registration of a reference facade to a newly acquired target image. We propose here to rely on semantic segmentation to improve the accuracy of that initial registration. Simultaneously we aim to improve the quality of the semantic segmentation through the registration. These two problems are jointly solved in a Expectation-Maximization framework. We especially introduce a bayesian model that use prior semantic segmentation as well as geometric structure of the facade reference modeled by $L_p$ Gaussian Mixtures. We show the advantages of our method in term of robustness to clutter and change of illumination on urban images from various database.