Dense outlier detection and open-set recognition based on training with noisy negative images

Documenti analoghi

• Weakly supervised training of universal visual concepts for multi-domain semantic segmentation
  di: Bevandić, Petra, et al.
  Pubblicazione: (2022)
• DEARLi: Decoupled Enhancement of Recognition and Localization for Semi-supervised Panoptic Segmentation
  di: Martinović, Ivan, et al.
  Pubblicazione: (2025)
• Outlier detection by ensembling uncertainty with negative objectness
  di: Delić, Anja, et al.
  Pubblicazione: (2024)
• Hybrid Open-set Segmentation with Synthetic Negative Data
  di: Grcić, Matej, et al.
  Pubblicazione: (2023)
• Sequential keypoint density estimator: an overlooked baseline of skeleton-based video anomaly detection
  di: Delić, Anja, et al.
  Pubblicazione: (2025)