SAFIRE is a large-scale benchmark for fire and smoke understanding in multimodal large language models (MLLMs), featuring 83,000 captioned images across 20 scenarios and 193,000 multiple-choice VQA questions derived from a 9.7K-image subset. The benchmark evaluates 10 dimensions of performance, from basic perception to higher-order reasoning, and employs a GPT‑5.4-assisted verification pipeline to ensure annotation quality. Experiments on ten open-source MLLMs (8B–38B) reveal an average accuracy of 61.9%, highlighting significant gaps in safety-critical reasoning, while fine-tuning vision encoders on just 7% of SAFIRE data boosts fire-scene classification accuracy from 20.1% to 64.5%. All resources are publicly available at https://risys-lab.github.io/SAFIRE/.
arXiv:2603. 29759v2 Announce Type: replace-cross Abstract: Recent advances in vision-language models (VLMs) have accelerated their application to indoor safety hazards assessment.
The paper investigates how evaluation protocols and ground‑truth quality affect the classification performance of Multimodal Large Language Models (MLLMs). It identifies and corrects key issues such as discarded out‑of‑list outputs, inflated distractor choices, and poor open‑world mapping, and shows that design choices like batch size, image ordering, and text encoder selection significantly influence accuracy. Using a multilabel reannotation of 625 ImageNet‑1k classes (ReGT), the study finds that corrected labels can boost MLLM performance by up to 10.8%, narrowing the gap with supervised models, and demonstrates that MLLMs can assist human annotators in about half of difficult cases.
The paper investigates whether state‑of‑the‑art multi‑modal large language models (LLMs) can detect deepfake images. Twelve recent LLMs were benchmarked against traditional detectors on multiple datasets, including real‑world deepfakes, and prompt tuning was used to improve performance. Results show that the best LLMs can match or exceed conventional methods on out‑of‑distribution data, though many LLMs perform poorly; larger model size helps, but newer versions and reasoning abilities do not necessarily improve detection.
SafeAtlas-VL introduces a large multimodal safety dataset with 1.5 million instances, rating image, request, and response risks on a five‑level ordinal scale across 15 harm categories and 55 subcategories. The accompanying SafeAtlas‑Bench provides 5,000 held‑out cases for evaluating ordinal predictions and continuous risk scores. Models trained on this data, including an 8B Guard model, achieve state‑of‑the‑art performance, outperforming prior benchmarks by about 4% in F1 score.
arXiv:2511.20022v3 Announce Type: replace-cross Abstract: Recent advancements in multimodal large language models (MLLMs) have shown strong understanding of driving scenes, drawing interest in their...
DisasterInsight is a building‑centric benchmark designed to evaluate vision‑language models (VLMs) for disaster response. Built on the xBD satellite dataset, it adds OpenStreetMap‑derived functional labels to 134,108 building instances and offers 15 task types, including instance assessment, scene counting, multi‑instance reasoning, and structured report generation. Experiments show that VLMs excel at visible damage detection but struggle with building function, multi‑instance reasoning, counting, and grounded reporting, and instruction tuning only partially mitigates these gaps.
arXiv:2606. 06217v1 Announce Type: cross Abstract: When a disaster unfolds, responders must answer not only what is happening, but also why it is happening, what will happen next, and what to do now, often from noisy low-altitude UAV views and under tight on-site compute constraints.
LADBench is a new benchmark comprising over 1,000 synthetic images that contain logical anomalies across four domains—Residential, Urban, Collaborative, and Nature. It introduces a Tiered Prompting Protocol that progressively discloses information to test how much explicit help a model needs to locate and reason about these faults. Evaluation of leading foundation models shows that even the best only reaches 70.11% accuracy, indicating that implicit logical fault detection remains largely unsolved.
arXiv:2601.18493v2 Announce Type: replace Abstract: Vision--language models (VLMs) show promise for disaster-response remote sensing, but existing benchmarks mainly emphasize scene-level or damage-ce...
arXiv:2609.05782v1 Announce Type: new Abstract: Vision-language models (VLMs) offer a promising alternative to conventional fire detection systems by reasoning about the semantic context of a scene a...
arXiv:2607. 06420v1 Announce Type: cross Abstract: Visual counting is a fundamental pillar of multimodal intelligence, requiring a seamless integration of fine-grained grounding and spatial reasoning.
arXiv:2606. 24759v1 Announce Type: cross Abstract: Recent multimodal large language models (MLLMs) have shown strong potential for autonomous driving scene understanding, yet existing methods still face a fundamental trade-off between temporal reasoning and spatial precision.