← Back to all news
arXiv Computer Vision

SigLIP2 for aerial fire risk classification

The paper presents an evaluation of the SigLIP2 image encoder for classifying fire risk from aerial images into seven categories. It introduces a reproducible split of the public FireRisk dataset and an implementation that tracks data provenance, preprocessing, and model selection. Two experiments—one with a frozen encoder probe and one with full model adaptation—show that full adaptation achieves 63.05% accuracy and 58.94% macro F1 on the validation set, outperforming the probe’s 55.95% accuracy and 50.19% macro F1.

Related stories

arXiv Machine Learning
Sep 16

A Sentinel-2 benchmark dataset for deep-learning active-fire segmentation across 25 California wildfires

The article introduces an open image dataset for active‑fire segmentation in satellite imagery, comprising 2,148 image‑mask pairs from 25 California wildfires captured between July 2020 and August 2026. Each 512×512 pixel, three‑channel image is a Sentinel‑2 Level‑2A composite of bands B12, B11, and B8A, with a fixed linear rendering applied uniformly. Masks distinguish background, short‑wave‑infrared rule‑based active fire, and invalid observations, and the dataset includes chip‑level metadata, an incident‑disjoint split, and a mask‑blind analyst review of 233 test chips.

By Shreyan Mitra, Mohammadreza Narimani, Parastoo Farajpoor
computer-visionbenchmarks
More like this →
arXiv AI
Sep 10

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

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/.

By Pengfei Li, Naufal Suryanto, Sicheng Zhang, Mohammad Alsharid, Muzammal Naseer
llmsmultimodalbenchmarkssafety
More like this →
arXiv Computer Vision
Sep 2

Scale-based Approach for Active Wildfire Segmentation on Satellite Imagery

The paper proposes a data-driven protocol that uses multispectral Landsat‑8 imagery and connected‑component analysis to characterize fire‑region size distributions for active wildfire segmentation. Three segmentation architectures—U‑Net, DeepLabV3+, and SegFormer—are evaluated under different SWIR‑based spectral configurations, with U‑Net showing the strongest robustness and SWIR2 consistently delivering the best results. The study highlights the importance of both spectral band selection and architectural design for robust satellite‑based active wildfire mapping, especially when training on low fire‑pixel density images.

By Matheus F. Kovaleski, Cristiano Premebida, Jo\~ao Ruivo Paulo
computer-vision
More like this →
arXiv Machine Learning
Oct 12

Shortcut Learning in Geospatial Foundation Model Embeddings for Cross-Disaster Damage Assessment

The study investigates shortcut learning in geospatial foundation model embeddings for cross-disaster damage assessment. By holding out entire disasters during evaluation, the authors demonstrate a significant drop in mean balanced accuracy across multiple datasets, revealing that models rely on event or scene characteristics rather than transferable evidence. They further show that simple feature-level transformations, such as supervised contrastive learning or PCA, can improve generalization to unseen disasters.

By Ali Shibli, Andrea Nascetti, Yifang Ban
rag
More like this →
arXiv AI
Sep 1

Toward Generalizable Deep Learning Based Peatland Fire Detection via Walsh Hadamard Transform and Domain Adaptation

arXiv:2603.02465v2 Announce Type: replace-cross Abstract: Machine learning-based wildfire detection has advanced significantly using deep learning models trained on large wildfire image and video dat...

By Emadeldeen Hamdan, Ahmad Faiz Tharima, Mohd Zahirasri Mohd Tohir, Dayang Nur Sakinah Musa, Erdem Koyuncu, Adam J. Watts, Ahmet Enis Cetin
fine-tuning
More like this →
arXiv AI
Aug 19

Training with synthetic data for drone detection in thermal imagery

The paper explores a synthetic-first training approach for detecting drones in medium- and long-wave infrared imagery, combining synthetic scene generation with fine-tuning on real data. It demonstrates that synthetic data can establish initial object representations, but real infrared data is crucial to close domain gaps and improve reliability. The study finds that aligning datasets has a greater impact on performance than increasing model size, and that semantic alignment in feature space is the strongest predictor of success, with radiometric factors like entropy and dynamic range also contributing.

By Tanel Liiv, Sander Soodla, Nzamba Bignoumba, Alma M. Liezenga, Toomas Pruuden
roboticsfine-tuningsafety
More like this →