This dataset contains 11027 labeled images for the detection of fire and smoke instances in diverse real-world scenarios. The annotations are provided in YOLO format with bounding boxes and class labels for two classes: fire and smoke. The dataset is divided into an 80% training set with 10,090 fire instances and 9724 smoke instances, a 10% Validation set with 1,255 fire and 1,241 smoke instances, and a 10% Test set with 1,255 fire and 1,241 smoke instances. This dataset is suitable for training and evaluating fire and smoke detection models, such as YOLOv8, YOLOv9, and similar deep learning-based frameworks in the context of emergency response, wildfire monitoring, and smart surveillance.

FASDD is a largest and most generalized Flame And Smoke Detection Dataset for object detection tasks, characterized by the utmost complexity in fire scenes, the highest heterogeneity in feature distribution, and the most significant variations in image size and shape. FASDD serves as a benchmark for developing advanced fire detection models, which can be deployed on watchtowers, drones, or satellites in a space-air-ground integrated observation network for collaborative fire warning. This endeavor provides valuable insights for government decision-making and fire rescue operations. FASDD contains fire, smoke, and confusing non-fire/non-smoke images acquired at different distances (near and far), different scenes (indoor and outdoor), different light intensities (day and night), and from various visual sensors (surveillance cameras, UAVs, and satellites). FASDD consists of three sub-datasets, a Computer Vision (CV) dataset (i.e. FASDD_CV), a Unmanned Aerial Vehicle (UAV) dataset (i.e. FASDD_UAV), and an Remote Sensing (RS) dataset (i.e. FASDD_RS). FASDD comprises 122,634 samples, with 70,581 annotated as positive samples and 52,073 labeled as negative samples. There are 113,154 instances of flame objects and 73,072 instances of smoke objects in the entire dataset. FASDD_CV contains 95,314 samples for general computer vision, while FASDD_UAV consists of 25,097 samples captured by UAV, and FASDD_RS comprises 2,223 samples from satellite imagery. FASDD_CV contains 73,297 fire instances and 53,080 smoke instances. The CV dataset exhibits considerable variation in image size, ranging from 78 to 10,600 pixels in width and 68 to 8,858 pixels in height. The aspect ratios of the images also vary significantly, ranging from 1:6.6 to 1:0.18. FASDD_UAV contains 36,308 fire instances and 17,222 smoke instances, with image aspect ratios primarily distributed between 4:3 and 16:9. In FASDD_RS, there are 2,770 smoke instances and 3,549 flame instances. The sizes of remote sensing images are predominantly around 1,000×1,000 pixels.FASDD is provided in three compressed files: FASDD_CV.zip, FASDD_UAV.zip, and FASDD_RS.zip, which correspond to the CV dataset, the UAV dataset, and the RS dataset, respectively. Additionally, there is a FASDD_RS_SWIR. zip folder storing pseudo-color images for detecting flame objects in remote sensing imagery. Each zip file contains two folders: "images" for storing the source data and "annotations" for storing the labels. The "annotations" folder consists of label files in four formats: YOLO, VOC, COCO, and TDML. The dataset is divided randomly into training, validation, and test sets, with a ratio of 1/2, 1/3, and 1/6, respectively, within each label format. In FASDD_CV, FASDD_UAV, and FASDD_RS, images and their corresponding annotation files have been individually sorted starting from 0. The flame and smoke objects in FASDD are given the labels "fire" and "smoke" for the object detection task, respectively. The names of all images and annotation files are prefixed with "Fire", "Smoke", "FireAndSmoke", and "NeitherFireNorSmoke", representing different categories for scene classification tasks.When using this dataset, please cite the following paper. Thank you very much for your support and cooperation：################################################################################使用数据集请引用对应论文，非常感谢您的关注和支持：Wang, M., Yue, P., Jiang, L., Yu, D., Tuo, T., & Li, J. (2025). An open flame and smoke detection dataset for deep learning in remote sensing based fire detection. Geo-spatial Information Science, 28(2), 511-526.################################################################################

Fire Detection Dataset - 85 videos

The dataset comprises 85 videos containing fire and smoke scenes, varying in length and content. Each video frame is annotated with bounding boxes that localize instances of fire and smoke, making it a comprehensive resource for fire monitoring, wildfire detection, and real-time monitoring systems. Designed to support detection systems, deep learning, and model training, this dataset is essential for improving fire management, early detection, and fire safety applications.

By leveraging this dataset, researchers and developers can enhance detection algorithms, train machine learning models, and improve accuracy in identifying fire events, smoke detection, and burnt areas under diverse environmental conditions.- Get the data

Each video sequence has been meticulously annotated by experts with precise bounding boxes that accurately localize all visible instances of fire and smoke throughout the frames.

💵 Buy the Dataset: This is a limited preview of the data. To access the full dataset, please contact us at https://unidata.pro to discuss your requirements and pricing options.

Researchers working on early warning systems will find the extensive footage of incipient fire stages especially valuable. Supports creation of fire prevention systems by analyzing fire behavior in different contexts.

🌐 UniData provides high-quality datasets, content moderation, data collection and annotation for your AI/ML projects

Here are a few use cases for this project:

1. Fire Emergency Identification: The model can be used in fire detection systems in public and private buildings. When it identifies fire or smoke and the presence of a human, it could trigger alarms and deploy necessary measures such as spraying fire retardant or auto-dialing emergency services.

2. Personal Safety Applications: In smart home systems, the model could provide real-time alerts to homeowners if fire or smoke is detected, especially if there's a human present, indicating potential danger.

3. Forest Fire Surveillance: The model can analyze drone or satellite imagery to identify forest fires and detect if anyone is trapped or injured within the vicinity, helping to strategize the response.

4. Industrial Safety: The model can be used in industries, particularly those with higher fire risk like oil and gas, chemical, and manufacturing, to monitor for fire or smoke and ensure the safety of the workers.

5. Disaster Response Training: The model can be used in simulations to train emergency response teams. For instance, the model would identify fire, smoke, and humans in various scenarios, providing realistic training opportunities for firefighters and rescue teams.

Fire And Smoke Detection 2

## Overview

Fire And Smoke Detection 2 is a dataset for object detection tasks - it contains Fire Smoke annotations for 5,512 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Smoke Fire Detection

## Overview

Smoke  Fire Detection is a dataset for object detection tasks - it contains Objects annotations for 11,689 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [BY-NC-SA 4.0 license](https://creativecommons.org/licenses/BY-NC-SA 4.0).

Fire Detection Dataset - 85 videos

The dataset comprises 85 videos containing fire and smoke scenes, varying in length and content. Each video frame is annotated with bounding boxes that localize instances of fire and smoke, making it a comprehensive resource for fire monitoring, wildfire detection, and real-time monitoring systems. Designed to support detection systems, deep learning, and model training, this dataset is essential for improving fire management, early detection, and… See the full description on the dataset page: https://huggingface.co/datasets/UniDataPro/fire-and-smoke-dataset.

The real-time capabilities of different models on fire-smoke detection.

Smoke and fire detection technology is a key technology for automatically realizing forest monitoring and forest fire warning. One of the most popular algorithms for object detection tasks is YOLOv5. However, it suffers from some challenges, such as high computational load and limited detection performance. This paper proposes a high-performance lightweight network model for detecting forest smoke and fire based on YOLOv5 to overcome these problems. C3Ghost and Ghost modules are introduced into the Backbone and Neck network to achieve the purpose of reducing network parameters and improving the feature’s expressing performance. Coordinate Attention (CA) module is introduced into the Backbone network to highlight the object’s important information about smoke and fire and to suppress irrelevant background information. In Neck network part, in order to distinguish the importance of different features in feature fusing process, the weight parameter of feature fusion is added which is based on PAN (path aggregation network) structure, which is named PAN-weight. Multiple sets of controlled experiments were conducted to confirm the proposed method’s performance. Compared with YOLOv5s, the proposed method reduced the model size and FLOPs by 44.75% and 47.46% respectively, while increased precision and mAP(mean average precision)@0.5 by 2.53% and 1.16% respectively. The experimental results demonstrated the usefulness and superiority of the proposed method. The core code and dataset required for the experiment are saved in this article at https://github.com/vinchole/zzzccc.git.

Quick Start Guide

Problem Type : Binary Classification Target Variable : Fire Alarm

Context

A smoke detector is a device that senses smoke, typically as an indicator of fire. Smoke detectors are usually housed in plastic enclosures, typically shaped like a disk about 150 millimetres (6 in) in diameter and 25 millimetres (1 in) thick, but shape and size vary.

--> Types of Smoke Detectors

1. Photoelectric Smoke Detector

A photoelectric smoke detector contains a source of infrared, visible, or ultraviolet light, a lens, and a photoelectric receiver. In some types, the light emitted by the light source passes through the air being tested and reaches the photosensor. The received light intensity will be reduced due to scattering from particles of smoke, air-borne dust, or other substances; the circuitry detects the light intensity and generates an alarm if it is below a specified threshold, potentially due to smoke. Such detectors are also known as optical detectors.

1. Ionization Smoke Detector

An ionization smoke detector uses a radioisotope to ionize air. If any smoke particles enter the open chamber, some of the ions will attach to the particles and not be available to carry the current in that chamber. An electronic circuit detects that a current difference has developed between the open and sealed chambers, and sounds the alarm

The author of this dataset has successfully created a smoke detection device with the help of IOT devices and AI model. (Check Acknowledgement )

About the dataset

Collection of training data is performed with the help of IOT devices since the goal is to develop a AI based smoke detector device. Many different environments and fire sources have to be sampled to ensure a good dataset for training. A short list of different scenarios which are captured:

• Normal indoor
• Normal outdoor
• Indoor wood fire, firefighter training area
• Indoor gas fire, firefighter training area
• Outdoor wood, coal, and gas grill
• Outdoor high humidity
• etc.

The dataset is nearly 60.000 readings long. The sample rate is 1Hz for all sensors. To keep track of the data, a UTC timestamp is added to every sensor reading.

Acknowledgement / Credits

The data is collected by Stefan Blattmann in his project Real-time Smoke Detection with AI-based Sensor Fusion. Author's GitHub : https://github.com/Blatts01

Forest Fire And Smoke Detection

## Overview

Forest Fire And Smoke Detection is a dataset for object detection tasks - it contains Fire annotations for 1,100 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

The results of smoke and fire detection based on different models.

Ensuring safety and safeguarding indoor properties require reliable fire detection methods. Traditional detection techniques that use smoke, heat, or fire sensors often fail due to false positives and slow response time. Existing deep learning-based object detectors fall short of improved accuracy in indoor settings and real-time tracking, considering the dynamic nature of fire and smoke. This study aimed to address these challenges in fire and smoke detection in indoor settings. It presents a hyperparameter-optimized YOLOv5 (HPO-YOLOv5) model optimized by a genetic algorithm. To cover all prospective scenarios, we created a novel dataset comprising indoor fire and smoke images. There are 5,000 images in the dataset, split into training, validation, and testing samples at a ratio of 80:10:10. It also used the Grad-CAM technique to provide visual explanations for model predictions, ensuring interpretability and transparency. This research combined YOLOv5 with DeepSORT (which uses deep learning features to improve the tracking of objects over time) to provide real-time monitoring of fire progression. Thus, it allows for the notification of actual fire hazards. With a mean average precision (mAP@0.5) of 92.1%, the HPO-YOLOv5 model outperformed state-of-the-art models, including Faster R-CNN, YOLOv5, YOLOv7 and YOLOv8. The proposed model achieved a 2.4% improvement in mAP@0.5 over the original YOLOv5 baseline model. The research has laid the foundation for future developments in fire hazard detection technology, a system that is dependable and effective in indoor scenarios.

The FIRESENSE database, developed within the FP7-ENV-244088
''FIRESENSE - Fire Detection and Management through a Multi-Sensor Network for
the Protection of Cultural Heritage Areas from the Risk of Fire and Extreme
Weather" project contains videos for testing flame and smoke detection algorithms.

Specifically:

a) for flame detection 11 positive and 16 negative videos are provided, while

b) for smoke detection, 13 positive and 9 negative videos are provided.

Results using this database are presented in many papers, including:

1. K. Dimitropoulos, P. Barmpoutis, N. Grammalidis, "Spatio-Temporal Flame Modeling and Dynamic Texture Analysis for Automatic Video-Based Fire Detection", IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), Vol. 25, No. 2, February 2015, pp. 339-351.

Ensuring safety and safeguarding indoor properties require reliable fire detection methods. Traditional detection techniques that use smoke, heat, or fire sensors often fail due to false positives and slow response time. Existing deep learning-based object detectors fall short of improved accuracy in indoor settings and real-time tracking, considering the dynamic nature of fire and smoke. This study aimed to address these challenges in fire and smoke detection in indoor settings. It presents a hyperparameter-optimized YOLOv5 (HPO-YOLOv5) model optimized by a genetic algorithm. To cover all prospective scenarios, we created a novel dataset comprising indoor fire and smoke images. There are 5,000 images in the dataset, split into training, validation, and testing samples at a ratio of 80:10:10. It also used the Grad-CAM technique to provide visual explanations for model predictions, ensuring interpretability and transparency. This research combined YOLOv5 with DeepSORT (which uses deep learning features to improve the tracking of objects over time) to provide real-time monitoring of fire progression. Thus, it allows for the notification of actual fire hazards. With a mean average precision (mAP@0.5) of 92.1%, the HPO-YOLOv5 model outperformed state-of-the-art models, including Faster R-CNN, YOLOv5, YOLOv7 and YOLOv8. The proposed model achieved a 2.4% improvement in mAP@0.5 over the original YOLOv5 baseline model. The research has laid the foundation for future developments in fire hazard detection technology, a system that is dependable and effective in indoor scenarios.

After coming across the open source data analysis tool created by Enigma, Bloomington FD (BFD) decided to pursue this data driven prevention project. After uploading 16 years of fire response data, the department was given a spreadsheet with our data analysis. The complexity of the project exceeded internal Fire Department capabilities and quickly moved to the collaborative project list established by our interdisciplinary team. This team based out of Indiana University includes data science researchers from the School of Informatics and Computing, leading technologists from the University’s Information Technology Services (UITS), and BFD members. The mapped data was created by Logan Paul, a graduate researcher in Prof. David Wild’s Integrative Data Science Laboratory and is much easier to use than the raw data. BFD plans to use the data to help focus smoke detector installations to areas that will have the biggest impact. If Bloomington's results are similar to other Cities across the nation, this data driven approach will increase our accuracy of smoke detector installations from 5-8 percent to nearly 65 percent. This represents a more efficient delivery of service that will also save lives. How it works: https://www.enigma.com/blog/developing-a-risk-model-for-residences-without-smoke-alarms Open Source algorithm: http://labs.enigma.io/smoke-signals/

Fire And Smoke Detection 1

## Overview

Fire And Smoke Detection 1 is a dataset for object detection tasks - it contains Fire annotations for 8,403 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Ensuring safety and safeguarding indoor properties require reliable fire detection methods. Traditional detection techniques that use smoke, heat, or fire sensors often fail due to false positives and slow response time. Existing deep learning-based object detectors fall short of improved accuracy in indoor settings and real-time tracking, considering the dynamic nature of fire and smoke. This study aimed to address these challenges in fire and smoke detection in indoor settings. It presents a hyperparameter-optimized YOLOv5 (HPO-YOLOv5) model optimized by a genetic algorithm. To cover all prospective scenarios, we created a novel dataset comprising indoor fire and smoke images. There are 5,000 images in the dataset, split into training, validation, and testing samples at a ratio of 80:10:10. It also used the Grad-CAM technique to provide visual explanations for model predictions, ensuring interpretability and transparency. This research combined YOLOv5 with DeepSORT (which uses deep learning features to improve the tracking of objects over time) to provide real-time monitoring of fire progression. Thus, it allows for the notification of actual fire hazards. With a mean average precision (mAP@0.5) of 92.1%, the HPO-YOLOv5 model outperformed state-of-the-art models, including Faster R-CNN, YOLOv5, YOLOv7 and YOLOv8. The proposed model achieved a 2.4% improvement in mAP@0.5 over the original YOLOv5 baseline model. The research has laid the foundation for future developments in fire hazard detection technology, a system that is dependable and effective in indoor scenarios.

1) Data Introduction • The Sensor-Fusion Smoke Detection Classification dataset is data related to smoke detectors that have reduced the number of fire victims.Fire can be prevented by the prediction results of smoke detectors.

2) Data Utilization (1) Sensor-Fusion Smoke Detection Classification data has characteristics that: • The dataset includes factors such as temperature, humidity, TVOC, CO2, H2 and Air Pressure. (2) Sensor-Fusion Smoke Detection Classification data can be used to: • Machine learning research: Help devise machine learning models that detect smoke and generate fire alarms with the help of IoT data.

Fire and smoke alarm manufacturers produce fire detection and alarm systems. Construction activity, especially nonresidential construction, primarily drives revenue since commercial, industrial and institutional buyers are the industry's largest markets. Manufacturers have combated volatile conditions through the current period, with the pandemic and climbing interest rates introducing unprecedented uncertainty. Even as critical markets recovered, supply chain disruptions, tariffs and low housing starts have prevented some manufacturers, especially smaller competitors, from fully recouping losses. Comparatively, the industry's leaders have displayed robust growth from stronger nonresidential markets, offering high-value solutions for core buyers. Overall, revenue has expanded at an estimated CAGR of 2.9% to $1.7 billion through the current period, including a 1.5% jump in 2025, where profit reached 14.2%. Companies have endured notable supply chain volatility through the current period, with various exogenous shocks and geopolitical events leading to shortages, longer lead times and other pricing concerns. Higher costs have pressured most fire and smoke alarm manufacturers, leading to elevated purchasing costs. Even so, leading producers were able to leverage brand reputations and long-term contracts to mitigate costs and expand profit. Manufacturers have also faced mounting pressure from imports, particularly from low-cost producers in Mexico, China and Malaysia. However, tariff policies may diminish import penetration while threatening supply chains. Regardless, leading fire alarm producers have managed to thrive in higher-end, custom-built markets, mitigating the threat posed by low-cost imports. Fire and smoke alarm manufacturers will benefit from a strong construction recovery and continued industrial growth through the outlook period. In particular, normalizing interest rates will encourage key industrial, commercial, institutional and residential clients to start new construction projects. Similarly, robust demand from data centers, driven by rapid AI integration and supportive government policies, will create additional, resilient revenue streams. Many companies will use excess returns to upgrade or replace existing systems, particularly as manufacturers rollout new safety, monitoring and connectivity features. Overall, revenue will climb at an expected CAGR of 2.3% to $1.9 billion through the outlook period, where profit will reach 14.6% of total revenue.