D
Danijel Skočaj
Researcher at University of Ljubljana
Publications - 85
Citations - 2529
Danijel Skočaj is an academic researcher from University of Ljubljana. The author has contributed to research in topics: Mobile robot & Anomaly detection. The author has an hindex of 22, co-authored 79 publications receiving 1754 citations. Previous affiliations of Danijel Skočaj include Bosch.
Papers
More filters
Journal ArticleDOI
Segmentation-based deep-learning approach for surface-defect detection
TL;DR: A segmentation-based deep-learning architecture that is designed for the detection and segmentation of surface anomalies and is demonstrated on a specific domain of surface-crack detection.
Journal ArticleDOI
Combining reconstructive and discriminative subspace methods for robust classification and regression by subsampling
TL;DR: This work presents a theoretical framework for achieving the best of both types of methods: an approach that combines the discrimination power of discriminative methods with the reconstruction property of reconstructive methods which enables one to work on subsets of pixels in images to efficiently detect and reject the outliers.
Journal ArticleDOI
Reconstruction by inpainting for visual anomaly detection
TL;DR: The RIAD approach (RIAD) randomly removes partial image regions and reconstructs the image from partial inpaintings, thus addressing the drawbacks of auto-enocoding methods.
Journal ArticleDOI
Deep Learning for Large-Scale Traffic-Sign Detection and Recognition
Domen Tabernik,Danijel Skočaj +1 more
TL;DR: A convolutional neural network approach, the mask R-CNN, is adopted to address the full pipeline of detection and recognition with automatic end-to-end learning, which is sufficient for deployment in practical applications of the traffic-sign inventory management.
Proceedings Article
Weighted and Robust Incremental Method for Subspace Learning
Danijel Skočaj,Ales Leonardis +1 more
TL;DR: An incremental method is presented, which sequentially updates the principal subspace considering weighted influence of individual images as well as individual pixels within an image, resulting in a novel incremental, weighted and robust method for subspace learning.