Hazard or Anomaly? Evaluating VLMs for Understanding Dangers and Discrepancies

• URL: http://arxiv.org/abs/2607.18325v1
• Date: Sat, 18 Jul 2026 02:24:06 GMT
• Title: Hazard or Anomaly? Evaluating VLMs for Understanding Dangers and Discrepancies
• Abstract summary: Vision-Language Models (VLMs) can interpret complex scenes and communicate safety-relevant information.<n>Current evaluations often frame danger recognition as a binary decision (Safe/Unsafe)<n>We introduce an explicit distinction between hazard and anomaly, and by separately recognizing hazardous and anomalous states.
• Score: 3.227001704699474
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Modern safety-critical systems increasingly rely on human-robot interaction to reduce disaster risk and support decision-making during emergencies. Vision-Language Models (VLMs) are promising for these settings because they can interpret complex scenes and communicate safety-relevant information, but they still require careful evaluation to ensure reliable safety reasoning. In particular, current evaluations often frame danger recognition as a binary decision (Safe/Unsafe), making it unclear whether a model is identifying true physical hazards or merely reacting to unusual scene elements. We address this limitation by introducing an explicit distinction between hazard and anomaly, and by separately recognizing hazardous and anomalous states. We evaluate several state-of-the-art VLMs across two datasets and multiple prompting strategies to test whether this distinction changes model behavior. Our results show that VLMs frequently misinterpret anomalousness as hazardousness, revealing an over-reliance on contextual irregularity as a proxy for danger. We further show that explicitly separating anomaly from hazard provides a more informative evaluation of VLM safety reasoning and exposes failure modes that binary safety judgments can obscure. Our public dataset is available on Roboflow https://app.roboflow.com/vlm-in-context-anomaly-and-hazard-detection/camera-ready-roman-ds.

Related papers

• When Seeing Overrides Knowing: Visual Dominance and Deferral-Based Method for Personalized Safety in VLMs [54.95635713318535]
  Vision-language models (VLMs) are increasingly deployed in high-stakes settings, where a response that is reasonable in general may still be unsafe for a particular user.<n>We study this problem of personalized safety in multimodal systems and introduce MPS-Bench, a benchmark of 5,181 scenarios from 584 real-world images.<n>We find that they almost always respond directly (86-99%) rather than seek missing context, and none exceeds 2.6/5 on personalized safety.<n>Motivated by this mechanism, we propose PRISM, a lightweight input monitor that uses bidirectional cross-modal modulation to predict when a query is likely to require
  arXiv  Detail & Related papers  (2026-09-03T03:58:56Z)
• EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety Guards [4.6400797556929545]
  Vision-language models (VLMs) are proposed as runtime safety guards for embodied agents in homes and factories.<n>We introduce EgoSafetyBench, an egocentric video benchmark of 1,200 robot-view scenarios annotated at half-second granularity.<n>We find that while guards reliably recognize videos containing hazards, they often miss specific hazardous moments.
  arXiv  Detail & Related papers  (2026-06-30T21:50:49Z)
• SafetyALFRED: Evaluating Safety-Conscious Planning of Multimodal Large Language Models [46.968844120076916]
  We introduce SafetyALFRED, built upon the embodied agent benchmark ALFRED, augmented with six categories of real-world kitchen hazards.<n>We evaluate eleven state-of-the-art models on not only hazard recognition, but also active risk mitigation through embodied planning.
  arXiv  Detail & Related papers  (2026-04-21T16:27:20Z)
• HazardArena: Evaluating Semantic Safety in Vision-Language-Action Models [87.35765363039638]
  We introduce HazardArena, a benchmark designed to evaluate semantic safety in Vision-Language-Action models.<n>We find that VLA models trained exclusively on safe scenarios often fail to behave safely when evaluated in their corresponding unsafe counterparts.<n>We propose a training-free Safety Option Layer that constrains action execution using semantic attributes or a vision-language judge.
  arXiv  Detail & Related papers  (2026-04-14T08:32:02Z)
• Uncertainty-Aware Vision-based Risk Object Identification via Conformal Risk Tube Prediction [7.840136094141999]
  Existing approaches make decisions and ignore uncertainty, which could lead to safety-critical failures.<n>Despite these challenges, current methods lack principled framework to model risk uncertainty jointly across space and time.<n>We propose Conformal Risk Prediction unified formulation that captures principled risk uncertainty, provides coverage guarantees for true risks, and produces calibrated risk scores with uncertainty estimates.
  arXiv  Detail & Related papers  (2026-03-25T04:13:42Z)
• SAVeS: Steering Safety Judgments in Vision-Language Models via Semantic Cues [50.928203120795786]
  We study whether multimodal safety behavior in vision-language models can be steered by simple semantic cues.<n>We introduce a semantic steering framework that applies controlled textual, visual, and cognitive interventions without changing the underlying scene content.
  arXiv  Detail & Related papers  (2026-03-19T16:18:00Z)
• Safety Not Found (404): Hidden Risks of LLM-Based Robotics Decision Making [12.400383981686801]
  One mistake by an AI system in a safety-critical setting can cost lives.<n>As Large Language Models (LLMs) become integral to robotics decision-making, the physical dimension of risk grows.<n>This paper addresses the urgent need to systematically evaluate LLM performance in scenarios where even minor errors are catastrophic.
  arXiv  Detail & Related papers  (2026-01-09T05:04:15Z)
• AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions [64.85086226439954]
  We present SAFE, a benchmark for assessing the safety of embodied VLM agents on hazardous instructions.<n> SAFE comprises three components: SAFE-THOR, SAFE-VERSE, and SAFE-DIAGNOSE.<n>We uncover systematic failures in translating hazard recognition into safe planning and execution.
  arXiv  Detail & Related papers  (2025-06-17T16:37:35Z)
• Better Safe Than Sorry? Overreaction Problem of Vision Language Models in Visual Emergency Recognition [12.054081112688074]
  Vision-Language Models (VLMs) have shown capabilities in interpreting visual content, but their reliability in safety-critical scenarios remains insufficiently explored.<n>We introduce VERI, a diagnostic benchmark comprising 200 synthetic images (100 contrastive pairs) and an additional 50 real-world images (25 pairs) for validation.<n>Each emergency scene is paired with a visually similar but safe counterpart through human verification.
  arXiv  Detail & Related papers  (2025-05-21T10:57:40Z)
• Towards a Multi-Agent Vision-Language System for Zero-Shot Novel Hazardous Object Detection for Autonomous Driving Safety [0.0]
  We propose a multimodal approach that integrates vision-language reasoning with zero-shot object detection.<n>We refine object detection by incorporating OpenAI's CLIP model to match predicted hazards with bounding box annotations.<n>Our findings highlight the strengths and limitations of current vision-language-based approaches.
  arXiv  Detail & Related papers  (2025-04-18T01:25:02Z)
• A Counterfactual Safety Margin Perspective on the Scoring of Autonomous Vehicles' Riskiness [52.27309191283943]
  This paper presents a data-driven framework for assessing the risk of different AVs' behaviors. We propose the notion of counterfactual safety margin, which represents the minimum deviation from nominal behavior that could cause a collision.
  arXiv  Detail & Related papers  (2023-08-02T09:48:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.