paper-with-me

Papers

Objectomaly: Objectness-Aware Refinement for OoD Segmentation with Structural Consistency and Boundary Precision

2025-07-10 · Jeonghoon Song, Sunghun Kim, Jaegyun Im, Byeongjoon Noh

Out-of-Distribution (OoD) segmentation is critical for safety-sensitive applications like autonomous driving. However, existing mask-based methods often suffer from boundary imprecision, inconsistent anomaly scores within objects, and false positives from background noise. We propose \textbf{\textit{Objectomaly}}, an objectness-aware refinement framework that incorporates object-level priors. Objectomaly consists of three stages: (1) Coarse Anomaly Scoring (CAS) using an existing OoD backbone, (2) Objectness-Aware Score Calibration (OASC) leveraging SAM-generated instance masks for object-level score normalization, and (3) Meticulous Boundary Precision (MBP) applying Laplacian filtering and Gaussian smoothing for contour refinement. Objectomaly achieves state-of-the-art performance on key OoD segmentation benchmarks, including SMIYC AnomalyTrack/ObstacleTrack and RoadAnomaly, improving both pixel-level (AuPRC up to 96.99, FPR$_{95}$ down to 0.07) and component-level (F1$-$score up to 83.44) metrics. Ablation studies and qualitative results on real-world driving videos further validate the robustness and generalizability of our method. Code will be released upon publication.

📄 PDF Abstract BibTeX arXiv:2507.07460

Code (0)

등록된 구현이 없습니다.

Tasks

Autonomous Driving

Similar Papers 제목 키워드 기반

FarSeg++: Foreground-Aware Relation Network for Geospatial Object Segmentation in High Spatial Resolution Remote Sensing Imagery

2023-07-13 · IEEE Transactions on Pattern Analysis and Machine Intelligence 2023 7 · Zhuo Zheng, Yanfei Zhong, Junjue Wang, Ailong Ma 외

Geospatial object segmentation, a fundamental Earth vision task, always suffers from scale variation, the larger intraclass variance of background, and foreground-background imbalance in high spatial resolution (HSR) rem…

Earth ObservationRelationRelation NetworkSegmentation+2

Mitigating Objectness Bias and Region-to-Text Misalignment for Open-Vocabulary Panoptic Segmentation

2026-03-22 · Nikolay Kormushev, Josip Šarić, Matej Kristan arxiv

Open-vocabulary panoptic segmentation remains hindered by two coupled issues: (i) mask selection bias, where objectness heads trained on closed vocabularies suppress masks of categories not observed in training, and (ii)…

Panoptic SegmentationImage Classification

Objectness-Aware Few-Shot Semantic Segmentation

2020-04-06 · Yinan Zhao, Brian Price, Scott Cohen, Danna Gurari

Few-shot semantic segmentation models aim to segment images after learning from only a few annotated examples. A key challenge for them is how to avoid overfitting because limited training data is available. While prior …

Few-Shot Semantic SegmentationSegmentationSemantic Segmentation

3D Guided Weakly Supervised Semantic Segmentation

2020-12-01 · Weixuan Sun, Jing Zhang, Nick Barnes

Pixel-wise clean annotation is necessary for fully-supervised semantic segmentation, which is laborious and expensive to obtain. In this paper, we propose a weakly supervised 2D semantic segmentation model by incorporati…

2D Semantic SegmentationSegmentationSemantic SegmentationWeakly supervised Semantic Segmentation+1

Turbulence-Robust Dynamic Object Segmentation with Multi-Signal Priors and SAM2 Refinement

2026-05-28 · Bolian Peng, Ying Tang, Xu Liu, Long Sun 외 arxiv

This technical report presents our solution for the CVPR 2026 UG2+ Challenge Track 3: Dynamic Object Segmentation in Turbulence (DOST). We design a training-free multi-signal segmentation pipeline that combines pretraine…

Object Segmentation