Constructing a Real-World Benchmark for Early Wildfire Detection with the New PYRONEAR-2025 Dataset

TL;DR

This work tackles the need for a scalable, real-world benchmark for early wildfire detection by introducing PyroNear_2025, a multi-source dataset with roughly 150,000 smoke-plume annotations on 50,000 images and 1,049 videos spanning four countries. It fuses in-house camera data, public networks, and synthetic imagery into two subsets: PyroNear_2025-I for single-frame detection and PyroNear_2025-V for video-based sequential modeling, enabling both lightweight and temporal approaches. Using a lightweight YOLOv8 baseline for images and a CNN‑LSTM sequence model for videos, the study shows the dataset is challenging (typical F1 around 0.70) but improves cross-dataset performance when combined with existing public data; video data further enables earlier detections and smoother predictions. The work provides open-source code and data to advance research and deployment of resource-efficient early wildfire detection systems in real-world, low-power settings.”

Abstract

Early wildfire detection (EWD) is of the utmost importance to enable rapid response efforts, and thus minimize the negative impacts of wildfire spreads. To this end, we present PYRONEAR-2025, a new dataset composed of both images and videos, allowing for the training and evaluation of smoke plume detection models, including sequential models. The data is sourced from: (i) web-scraped videos of wildfires from public networks of cameras for wildfire detection in-the-wild, (ii) videos from our in-house network of cameras, and (iii) a small portion of synthetic and real images. This dataset includes around 150,000 manual annotations on 50,000 images, covering 640 wildfires, PYRONEAR-2025 surpasses existing datasets in size and diversity. It includes data from France, Spain, Chile and the United States. Finally, it is composed of both images and videos, allowing for the training and evaluation of smoke plume detection models, including sequential models. We ran cross-dataset experiments using a lightweight state-of-the-art object detection model, as the ones used in-real-life, and found out the proposed dataset is particularly challenging, with F1 score of around 70\%, but more stable than existing datasets. Finally, its use in concordance with other public datasets helps to reach higher results overall. Last but not least, the video part of the dataset can be used to train a lightweight sequential model, improving global recall while maintaining precision for earlier detections. [We make both our code and data available online](https://github.com/joseg20/wildfires2025).

Figures (9)

• Figure 1: Examples of our dataset, containing real images and videos from France, Spain, United States and Chile, and synthetic images.
• Figure 2: Summary of the whole process to create the video and the image datasets. ALERTWildfire data was collected from the web, FigLib data (without bounding box labels) comes from a published research paper, PyroNear data come from in-house data that we collected, Synthetic images were generated.
• Figure 3: Snapshot of the Smoke Plume Annotation Platform.
• Figure 4: F1 scores obtained by training on each dataset (x-axis) and evaluating on all others, including self-evaluation.
• Figure 5: Comparison of the predictions of a single-frame model versus a video model on a set of videos from FigLib.