SKLFS-WildFire2025 Test
Overview
SKLFS-WildFire2025 Test is a large-scale forest smoke detection test set introduced in LwST-FSD: Efficient Forest Smoke Detection via Lightweight Spatio-temporal Modeling. It is designed as an evaluation benchmark for smoke detection in real-world surveillance scenarios.
The dataset contains 8,363 video clips, totaling 152,700 annotated frames. Among them, 48,777 smoke-affected images are extracted from 3,952 video segments. All samples are reserved exclusively for testing, providing a large and diverse benchmark for evaluating forest smoke detection models.
Key Features
- Real-world forest smoke detection test set.
- 8,363 video clips.
- 152,700 annotated frames.
- 48,777 smoke-affected images from 3,952 video segments.
- Reserved exclusively for testing.
- Designed to evaluate smoke detection under large-scale and diverse real-world conditions.
Citation
If you use this dataset, please cite:
@article{WANG2025122481,
title = {LwST-FSD: Efficient forest smoke detection via lightweight spatio-temporal modeling},
journal = {Information Sciences},
volume = {719},
pages = {122481},
year = {2025},
issn = {0020-0255},
doi = {https://doi.org/10.1016/j.ins.2025.122481},
author = {Chong Wang and Hai Song and Zhong Wang and Zhilin Shan and Qixing Zhang},
}