SigLIP2 for aerial fire risk classification

Yunus Serhat Bıçakçı

cs.CVOct 2, 2026

SigLIP2 for aerial fire risk classification

Authors: Yunus Serhat Bıçakçı

Organizations: Department of Artificial Intelligence and Machine Learning Faculty of Applied Sciences, Marmara University Istanbul, Türkiye

Abstract

We examine the transfer of a pretrained SigLIP2 image encoder to seven class fire risk classification from aerial imagery. We introduce a reproducible partition of the public FireRisk training mirror and an implementation that records data provenance, preprocessing and model selection. Two initial runs compare a frozen encoder probe with full model adaptation. On the validation partition, full adaptation reaches 63.05% accuracy and 58.94% macro F1, compared with 55.95% and 50.19% for the probe. Both runs use one training seed and select their checkpoint on the same validation partition. These development results support further evaluation of SigLIP2 but do not establish performance on an independent test set or unseen regions. The accompanying code provides a common framework for repeated experiments and comparisons with additional visual encoders.

Figures & tables

Figure 2

Recipe	Trainable	Accuracy	Macro F1	Time	Memory
	parameters	(%)	(%)	(min)	(GiB)
Frozen encoder probe	6,919	55.95	50.19	12.1	0.74
Full adaptation	92,891,143	63.05	58.94	19.0	1.96

Table 3

Class	Support	F1 (%)		Recall (%)
		Frozen	Full	Frozen	Full
Very low	3,264	60.86	69.87	59.56	68.96
Low	1,606	35.35	41.76	36.80	44.83
Moderate	1,293	29.40	42.03	25.91	42.00
High	945	40.27	46.53	46.56	45.08
Very high	490	21.29	39.90	14.49	33.47

Table 4

Image Classification

Image classification assigns semantic labels to images, enabling recognition of objects, scenes, conditions, and fine-grained categories.

See all 386 papers

SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs
Sep 7, 2026Pengfei Li, Naufal Suryanto, Sicheng Zhang +2LLM Safety Benchmarks LLM Safety Evaluation
Khalifa University · University of Western Australia
HumP-KD: A Hybrid Uncertainty-Aware Multi-Stage Progressive Knowledge Distillation Framework for Efficient Fire Classification
Jun 12, 2026Mohammed Arif Mainuddin, Najifa Tabassum, Omar Ibne Shahid +1Efficient Neural Network Inference Cross-Architecture Knowledge Distillation
Electrical and Computer Engineering North South University Dhaka, Bangladesh
Representation Risk in Pretrained Image Encoders
Sep 28, 2026Ardyn Nordstrom, Morgan Nordstrom, Vamuyan Sesay +1Vision Foundation Models Predictive Representation Learning
School of Public Policy and Administration, Carleton University · Independent Researcher · Department of Economics, Carleton University

Sep 7, 2026J/K move · Enter open · S save

SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs

Pengfei Li, Naufal Suryanto, Sicheng Zhang, Mohammad Alsharid+1

Khalifa University · University of Western Australia

LLM Safety Benchmarks LLM Safety Evaluation Multimodal Model Evaluation

Open paper AskPDF arXiv

Multimodal Large Language Models (MLLMs) show strong progress on vision-language tasks, yet their reliability in safety-critical settings remains underexplored. Fire-smoke understanding is central to public safety and disaster response, but most existing benchmarks lack diverse real-world scenarios and context-aware evaluation. We introduce SAFIRE, a large-scale benchmark for fire-smoke understanding in MLLMs, comprising 83K captioned images from 20 scenarios and 193K multiple-choice VQA (MCVQA) generated from a 9.7K-image subset, spanning 10 evaluation dimensions from basic perception to higher-order reasoning. A GPT-5.4-assisted multi-stage verification pipeline with MLLM majority voting ensures annotation quality. Evaluating ten open-source MLLMs (8B-38B) yields an average accuracy of 61.9%, exposing major gaps in safety-critical reasoning. We further show that adapting vision encoders with only 7% of our domain-specific data boosts fire-scene classification accuracy from 20.1% to 64.5%, indicating that carefully curated data can yield substantial gains even when data volume is limited. All datasets, models, and code are available at https://risys-lab.github.io/SAFIRE/.

Abstract

Image Classification

Explore similar work

SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs

HumP-KD: A Hybrid Uncertainty-Aware Multi-Stage Progressive Knowledge Distillation Framework for Efficient Fire Classification

Representation Risk in Pretrained Image Encoders