Pérez-García de la Puente, Natalia Lourdesdel Amor, RocíoGarcía-Torres, FernandoColomer, AdriánNaranjo Ornedo, Valeriana2024-04-302024-04-302023-11-24978-3-031-48232-8https://riunet.upv.es/handle/10251/203857[EN] Artificial Intelligence (AI) provides a fundamental aid in building operations, allowing infrastructure inspection and compliance with safety standards. In the collaborative tasks involved, detecting areas of interest, such as surface defects, is crucial. A drawback of supervised AI-based approaches is that they require manual annotation, which entails additional costs. This paper presents a novel unsupervised anomaly detection approach for locating defects based on generative models that learn the distribution of defect-free images. Using attention maps to validate in a subset, we propose a formulation that does not require accessing labelled images, enabling task automation, maintenance optimisation and cost reduction.Reserva de todos los derechosVisual InspectionInfrastructure InspectionDefectsUnsupervised SegmentationTEORÍA DE LA SEÑAL Y COMUNICACIONESESTADISTICA E INVESTIGACION OPERATIVAComunicación en congreso10.1007/978-3-031-48232-8_14Abierto