Lune

CVPR2026Top-tier venue

ClimaOoD: Improving Anomaly Segmentation via Physically Realistic Synthetic Data

Yuxing Liu, Zheng Li, Huanhuan Liang, Ji Zhang, Zeyu Sun, Yong Liu

Abstract

Anomaly segmentation seeks to detect and localize unknown or out-of-distribution (OoD) objects that fall outside predefined semantic classes a capability essential for safe autonomous driving. However, the scarcity and limited diversity of anomaly data severely constrain model generalization in open-world environments. Existing approaches mitigate this issue through synthetic data generation, either by copy-pasting external objects into driving scenes or by leveraging text-to-image diffusion models to inpaint anomalous regions. While these methods improve anomaly diversity, they often lack contextual coherence and physical realism, resulting in domain gaps between synthetic and real data. In this paper, we present ClimaDrive, a semantics-guided image-to-image framework for synthesizing semantically coherent, weather-diverse, and physically plausible OoD driving data. ClimaDrive unifies structure-guided multi-weather generation with prompt-driven anomaly inpainting, enabling the creation of visually realistic training data. Based on this framework, we construct ClimaOoD, a large-scale benchmark spanning six representative driving scenarios under both clear and adverse weather conditions. Extensive experiments on four state-of-the-art methods show that training with ClimaOoD leads to robust improvements in anomaly segmentation. Across all methods, AUROC, AP, and FPR95 show notable gains, with FPR95 dropping from 3.97 to 3.52 for RbA on Fishyscapes LAF. These results demonstrate that ClimaOoD enhances model robustness, offering valuable training data for better generalization in open-world anomaly detection.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Entropy Maximization and Meta Classification for Out-of-Distribution Detection in Semantic Segmentation cites Detecting the Unexpected via Image Resynthesis
  • RbA: Segmenting Unknown Regions Rejected by All cites Entropy Maximization and Meta Classification for Out-of-Distribution Detection in Semantic Segmentation
  • RbA: Segmenting Unknown Regions Rejected by All cites Detecting the Unexpected via Image Resynthesis
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on17

  1. High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
  2. BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and GenerationJunnan Li, Dongxu Li, Caiming Xiong, Steven C. H. HoiICML 2022 · 6,549 citations
  3. Detecting the Unexpected via Image ResynthesisKrzysztof Lis, Krishna Kanth Nakka, Pascal Fua, Mathieu SalzmannICCV 2019 · 217 citations
  4. Entropy Maximization and Meta Classification for Out-of-Distribution Detection in Semantic SegmentationRobin Chan, Matthias Rottmann, Hanno GottschalkICCV 2021 · 200 citations
  5. RbA: Segmenting Unknown Regions Rejected by AllNazir Nayal, Misra Yavuz, João F. Henriques, Fatma GüneyICCV 2023 · 73 citations

Related papers

  1. Beyond Pixel Uncertainty: Bounding the OoD Objects in Road ScenesHuachao Zhu, Zelong Liu, Zhichao Sun, Yuda Zou et al.ICCV 2025 · 1 citation
  2. Scaling Out-of-Distribution Detection for Real-World SettingsDan Hendrycks, Steven Basart, Mantas Mazeika, Andy Zou et al.ICML 2022 · 653 citations
  3. Segmenting Objectiveness and Task-awareness Unknown Region for Autonomous DrivingMi Zheng, Guanglei Yang, Zitong Huang, Zhenhua Guo et al.ACM MM 2025 · 1 citation
  4. Improving Anomaly Segmentation with Multi-Granularity Cross-Domain AlignmentJi Zhang, Xiao Wu, Zhi-Qi Cheng, Qi He et al.ACM MM 2023 · 7 citations
  5. Segment Every Out-of-Distribution ObjectWenjie Zhao, Jia Li, Xin Dong, Yu Xiang et al.CVPR 2024 · 11 citations
Dusk over the sea between two cliffs drawn in fine vertical lines