Zobrazeno 1 - 10
of 276
pro vyhledávání: '"Šegvić, Siniša"'
Backdoor attacks change a small portion of training data by introducing hand-crafted triggers and rewiring the corresponding labels towards a desired target class. Training on such data injects a backdoor which causes malicious inference in selected
Externí odkaz:
http://arxiv.org/abs/2409.01185
Domain adaptive panoptic segmentation promises to resolve the long tail of corner cases in natural scene understanding. Previous state of the art addresses this problem with cross-task consistency, careful system-level optimization and heuristic impr
Externí odkaz:
http://arxiv.org/abs/2407.14110
Outlier detection is an essential capability in safety-critical applications of supervised visual recognition. Most of the existing methods deliver best results by encouraging standard closed-set models to produce low-confidence predictions in negati
Externí odkaz:
http://arxiv.org/abs/2402.15374
Discriminative learning effectively predicts true object class for image classification. However, it often results in false positives for outliers, posing critical concerns in applications like autonomous driving and video surveillance systems. Previ
Externí odkaz:
http://arxiv.org/abs/2310.06085
Most approaches to dense anomaly detection rely on generative modeling or on discriminative methods that train with negative data. We consider a recent hybrid method that optimizes the same shared representation according to cross-entropy of the disc
Externí odkaz:
http://arxiv.org/abs/2305.15227
Autor:
Schubert, Marius, Riedlinger, Tobias, Kahl, Karsten, Kröll, Daniel, Schoenen, Sebastian, Šegvić, Siniša, Rottmann, Matthias
Labeling datasets for supervised object detection is a dull and time-consuming task. Errors can be easily introduced during annotation and overlooked during review, yielding inaccurate benchmarks and performance degradation of deep neural networks tr
Externí odkaz:
http://arxiv.org/abs/2303.06999
Real-world deployment of reliable object detectors is crucial for applications such as autonomous driving. However, general-purpose object detectors like Faster R-CNN are prone to providing overconfident predictions for outlier objects. Recent outlie
Externí odkaz:
http://arxiv.org/abs/2302.07106
Autor:
Grcić, Matej, Šegvić, Siniša
Open-set segmentation can be conceived by complementing closed-set classification with anomaly detection. Many of the existing dense anomaly detectors operate through generative modelling of regular data or by discriminating with respect to negative
Externí odkaz:
http://arxiv.org/abs/2301.08555
Most dense recognition approaches bring a separate decision in each particular pixel. These approaches deliver competitive performance in usual closed-set setups. However, important applications in the wild typically require strong performance in pre
Externí odkaz:
http://arxiv.org/abs/2301.03407
Publikováno v:
International Journal of Computer Vision, 2024, 1-23
Deep supervised models have an unprecedented capacity to absorb large quantities of training data. Hence, training on multiple datasets becomes a method of choice towards strong generalization in usual scenes and graceful performance degradation in e
Externí odkaz:
http://arxiv.org/abs/2212.10340