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

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Song, Jeonghoon, Kim, Sunghun, Im, Jaegyun, Noh, Byeongjoon
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908445724114944
author Song, Jeonghoon
Kim, Sunghun
Im, Jaegyun
Noh, Byeongjoon
author_facet Song, Jeonghoon
Kim, Sunghun
Im, Jaegyun
Noh, Byeongjoon
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2507_07460
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Objectomaly: Objectness-Aware Refinement for OoD Segmentation with Structural Consistency and Boundary Precision
Song, Jeonghoon
Kim, Sunghun
Im, Jaegyun
Noh, Byeongjoon
Computer Vision and Pattern Recognition
Artificial Intelligence
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.
title Objectomaly: Objectness-Aware Refinement for OoD Segmentation with Structural Consistency and Boundary Precision
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2507.07460