相关参考文献
注意：仅列出部分参考文献，下载原文获取全部文献信息。
Article
Computer Science, Artificial Intelligence
Kai Han et al.
Summary: Transformer, a deep neural network with a self-attention mechanism, has been initially used in natural language processing and is now gaining attention in computer vision tasks. Transformer-based models perform as well as or even better than convolutional and recurrent neural networks in various visual benchmarks. This paper reviews vision transformer models, categorizes them based on different tasks, and analyzes their advantages and disadvantages. The discussed categories include backbone network, high/mid-level vision, low-level vision, and video processing. Efficient methods for applying transformer in real device-based applications are also explored. The challenges and further research directions for vision transformers are discussed as well.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Review
Medicine, General & Internal
Adnane Ait Nasser et al.
Summary: Chest X-ray radiography (CXR) is a widely used medical imaging modality for detecting life-threatening diseases. Computer-aided detection (CAD) using machine learning (ML) and deep learning (DL) techniques has shown great potential in improving diagnosis efficiency. This review provides a detailed overview of publicly available CXR datasets, deep learning models, and techniques used for image preprocessing, as well as discusses challenges and the importance of interpretability and explainability in DL models.
Article
Chemistry, Analytical
Goncalo Perrolas et al.
Summary: This study proposes a new method for segmenting fire and smoke regions in high resolution images using a multi-resolution iterative quad-tree search algorithm. The method improves accuracy and efficiently handles small regions in the early stages of fire ignition.
Review
Engineering, Electrical & Electronic
Abdelmalek Bouguettaya et al.
Summary: The study focused on early detection of wildfires in forest and wildland areas using deep learning-based computer vision algorithms to prevent and reduce disastrous losses in terms of human lives and forest resources.
Article
Construction & Building Technology
James Pincott et al.
Summary: Current fire detection technology has limitations in providing accurate information for fire services. This study aims to develop a vision-based fire and smoke detection system using deep learning techniques, showing potential applications in indoor spaces.
BUILDING SERVICES ENGINEERING RESEARCH & TECHNOLOGY
(2022)
Article
Geography, Physical
Shubhangi Chaturvedi et al.
Summary: Early stage smoke detection using image and video analysis in the outdoor environment is crucial due to its wide applications in fire prevention and environmental safety. Various techniques, including conventional image processing, machine learning, and deep learning, have been proposed for real-time smoke detection. Key characteristics such as smoke pattern, motion analysis, color, and texture are utilized to identify smoke in the outdoor environment.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2022)
Article
Computer Science, Information Systems
Rajib Ghosh et al.
Summary: This article proposes a novel hybrid deep learning model that combines CNN and RNN for feature extraction to detect forest fire. The performance of the proposed system has been evaluated on two publicly available datasets, showing very high classification accuracy.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Chemistry, Analytical
Rafik Ghali et al.
Summary: This article introduces the use of Deep Learning and vision transformers to detect and classify wildfires. A novel deep ensemble learning method combining EfficientNet-B5 and DenseNet-201 models is proposed for wildfire identification and classification using aerial images. Additionally, TransUNet and TransFire models along with EfficientSeg are utilized for wildfire region segmentation and precise fire source determination. The results obtained demonstrate the efficiency of these approaches for wildfire classification and segmentation.
Article
Environmental Sciences
Houda Harkat et al.
Summary: This study addresses the issue of expensive wildfire detection using IoT sensors by implementing an on-site detection system that detects fire pixels in real-time. By deploying the Deeplabv3+ deep learning architecture and image processing techniques, the accuracy and efficiency of fire detection are significantly improved.
Article
Green & Sustainable Science & Technology
Lei Zhao et al.
Summary: The improved Fire-YOLO deep learning algorithm effectively detects small targets, fire-like and smoke-like targets in forest fire images, and achieves real-time fire detection under different natural lights, enhancing network performance.
Article
Forestry
Qilin Xue et al.
Summary: This paper examines the classification and identification of forest fires and proposes an improved model based on deep learning. By optimizing the loss function and introducing attention mechanisms, the model can accurately recognize different types of fires in complex backgrounds. Experimental results show significant improvements in detection performance compared to traditional methods.
Article
Forestry
Ziqi Wang et al.
Summary: Forest fires have caused significant damage worldwide. Fully convolutional network architectures have shown promise in segmenting flames for fire monitoring. However, research on applying these models in forest fire scenarios is limited. This study constructs a dataset of remote-sensing images captured by unmanned aerial vehicles (UAVs) and selects four semantic segmentation models and two backbone networks for testing and analysis. The results show that the U-Net model with Resnet50 as a backbone network has the highest segmentation accuracy, while the DeepLabV3+ model with Resnet50 provides faster real-time performance. FCN and PSPNet have poor segmentation performance and are not suitable for forest fire detection scenarios.
Article
Forestry
Lin Zhang et al.
Summary: In this paper, a FT-ResNet50 model based on transfer learning is proposed for forest fire identification based on UAV images. The experimental results show that the model achieves better accuracy in fire recognition compared to baseline models.
Article
Environmental Sciences
Zhihao Guan et al.
Summary: This paper presents two deep learning methods for forest-fire monitoring: an image classification method based on channel domain attention mechanism with an accuracy of 93.65%, and a novel instance segmentation method (MaskSU R-CNN) based on MS R-CNN model with precision of 91.85%, recall of 88.81%, F1-score of 90.30%, and mIoU of 82.31%. Our method achieves satisfactory results on forest-fire dataset compared to many state-of-the-art segmentation models.
Article
Mathematical & Computational Biology
Varanasi L. V. S. K. B. Kasyap et al.
Summary: With the increase in deforestation and global warming, forest fires have become more frequent. This article proposes the use of deep learning techniques and unmanned aerial vehicles for forest fire prediction and management. The proposed model outperforms traditional methods in fire detection and prediction.
COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE
(2022)
Article
Computer Science, Artificial Intelligence
Sengul Dogan et al.
Summary: This work develops an accurate fire warning model using images, utilizing two new deep feature engineering models and pretrained ResNet networks for feature extraction. Support vector machine classifiers and ensemble models are used for classification. The developed models achieve high classification accuracy and further testing on larger databases is recommended.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Editorial Material
Chemistry, Analytical
Yun Zhang et al.
Article
Forestry
Zhenyang Xue et al.
Summary: This paper proposes an improved forest fire small-target detection model based on YOLOv5, which enhances the detection ability of small forest fire targets by improving the model structure and adding attention modules.
Article
Forestry
Ji Lin et al.
Summary: This paper proposes an improved small-target forest fire detection model, STPM_SAHI, which utilizes the Swin Transformer backbone network and PAFPN feature fusion network to enhance the sensitivity and accuracy of forest fire detection. By integrating the Slicing Aided Hyper Inference technology, it solves the difficulty in detecting small-target forest fires and significantly improves the detection accuracy.
Proceedings Paper
Computer Science, Artificial Intelligence
Jiaxun Cui et al.
Summary: Optical sensors and learning algorithms for autonomous vehicles have advanced significantly, but their reliability is limited by line-of-sight sensing and the brittleness of data-driven methods. Cooperative perception between vehicles can enhance autonomous driving in dangerous or emergency situations.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Paul Hongsuck Seo et al.
Summary: The paper introduces a new video pretraining framework, MV-GPT, which has the ability to generate sentences and is used for multimodal video captioning tasks. By training a multimodal video encoder and a sentence decoder jointly and leveraging future utterances as additional text sources, a bidirectional generation objective is proposed to overcome the lack of captions in unlabelled videos. The model achieves state-of-the-art performance on multiple benchmarks.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Yulin Wang et al.
Summary: This paper presents an improved video recognition method that reduces spatial redundancy and introduces differentiable operations for better computational efficiency and simpler training process. Extensive experiments on multiple datasets demonstrate its effectiveness.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
(2022)
Review
Oncology
Ella Mahoro et al.
Summary: This review explores the application of deep learning in medical imaging research, particularly in early breast cancer screening. The review summarizes deep learning methods, data availability, and various screening methods for breast cancer, and discusses the limitations and opportunities of integrating artificial intelligence into breast cancer clinical practice.
Article
Computer Science, Information Systems
Xiwen Chen et al.
Summary: Current forest monitoring technologies have limitations in detecting and assessing wildfires, but drones are a valuable tool for early detection and assessment in remote forests. Limited aerial datasets have hindered research in data-driven fire detection and modeling techniques. This study presents a multi-modal UAV-collected dataset and a deep learning-based methodology for fire detection.
Article
Chemistry, Multidisciplinary
Shaoxiong Zheng et al.
Summary: An improved dynamic convolutional neural network (DCNN) model, named DCN_Fire, was established based on the traditional DCNN model for accurately identifying the risk of a forest fire. Transfer learning and principal component analysis were used to enhance the model's accuracy and speed. The results showed that the improved DCNN model had excellent recognition speed and accuracy, providing a technical reference for preventing and tackling forest fires.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Civil
James McCoy et al.
Summary: This paper proposes a novel approach for detecting unauthorized or malicious UAVs using a hybrid feature set of acoustic, image/video, and wireless RF signals, outperforming existing methods in terms of detecting malicious UAVs.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Multidisciplinary
Karol Bot et al.
Summary: This paper provides a review of recent applications of machine learning methods for wildfire management decision support, summarizing these applications based on case study type, machine learning method, case study location, and performance metrics. It is concluded that the adoption of machine learning methods can enhance support in different fire management phases.
Article
Engineering, Multidisciplinary
S. Geetha et al.
Summary: The risk of fires in urban buildings is increasing, and traditional methods of fire detection using smoke sensors have limitations. The introduction of video surveillance systems has provided new opportunities for identifying smoke and flame from a distance. Various methods, including image processing algorithms and CNNs, have been proposed to address the challenges of processing large amounts of data and improve fire and smoke detection in videos and images.
Article
Engineering, Multidisciplinary
Han-Soo Choi et al.
Summary: This paper introduces a semantic fire image segmentation method using a convolutional neural network with a stronger feedback effect achieved through the middle skip connection. The proposed model outperforms previous fire segmentation deep learning models and image processing algorithms in various evaluation metrics.
Article
Multidisciplinary Sciences
Vladimir Sergeevich Bochkov et al.
Summary: This article presents an AI-based solution for multiclass fire segmentation and develops a custom dataset for research purposes. By comparing different configurations of UNet models and choosing the best main metric for the loss function, new combinational methods to address the problem of detecting flame regions at boundaries were introduced. UUNet-concatenative and wUUNet models were proposed, showing significant improvements and considered state-of-the-art.
Article
Computer Science, Artificial Intelligence
Kyungmin Song et al.
Summary: The study proposes a binary semantic segmentation method for fire image prediction using deep learning, aiming to achieve fast prediction on embedded devices. By employing techniques such as depthwise separable convolution, the network parameters and model size were significantly reduced while enhancing prediction performance.
MACHINE VISION AND APPLICATIONS
(2021)
Article
Forestry
Renjie Xu et al.
Summary: The text discusses the challenges of forest fire detection and the application of deep learning technology. A novel ensemble learning method is proposed to improve detection accuracy and reduce false positives by integrating multiple learners.
Article
Computer Science, Artificial Intelligence
Shengying Wang et al.
Summary: This study proposes a deep learning fire recognition algorithm based on model compression and lightweight requirements, utilizing the MobileNetV3 model to simplify the conventional network structure and improve detection accuracy through knowledge distillation. Experimental results demonstrate significant advantages in reducing model parameters and inference time compared to existing algorithms.
JOURNAL OF REAL-TIME IMAGE PROCESSING
(2021)
Article
Computer Science, Hardware & Architecture
Alireza Shamsoshoara et al.
Summary: Wildfires in the US are highly destructive and dangerous, highlighting the need for technological advancements to minimize risks. The FLAME dataset and methods offer valuable tools for firefighters and researchers to optimize fire management strategies. By utilizing machine learning solutions, such as binary classification and fire detection segmentation, significant progress has been made in accurately identifying and managing fires.
Article
Forestry
Jin Pan et al.
Summary: Forest fires are catastrophic events that have global impact. This paper introduces a novel framework for collaborative region detection and grading of fire smoke using weakly supervised fine segmentation and Faster R-CNN. The proposed framework can perform early-stage alarm, region detection, classification, and grading of forest fire smoke simultaneously.
Article
Environmental Sciences
Rafik Ghali et al.
Summary: This study explores the potential of using vision Transformers for forest fire segmentation, presenting two frameworks based on Transformers that outperform current methods. Extensive evaluations showed superior performance, achieving state-of-the-art results in forest fire pixel mis-classification reduction and finer detection of fire shape.
Proceedings Paper
Computer Science, Cybernetics
Rafik Ghali et al.
Summary: The study proposed three deep convolutional networks for segmenting forest fire pixels and detecting fire areas, challenging the limitations of fire detection methods by varying loss functions. The three models showed excellent performance in terms of accuracy and F1-score, reliably segmenting fire pixels and detecting the shape of forest fire areas.
2021 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS (SMC)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Milad Niknejad et al.
Summary: The study introduced a Convolutional Neural Network (CNN) for joint classification and segmentation of fire in images, utilizing spatial self-attention and channel attention to improve fire segmentation performance.
20TH IEEE INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS (ICMLA 2021)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Jeya Maria Jose Valanarasu et al.
Summary: Deep convolutional neural networks have been widely adopted in medical image segmentation, but lack understanding of long-range dependencies due to inherent biases in convolutional architectures. Transformer-based architectures leverage self-attention mechanism to encode long-range dependencies, motivating the exploration of transformer solutions for medical image segmentation tasks.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT I
(2021)
Article
Computer Science, Interdisciplinary Applications
Tran Minh Quan et al.
Summary: Cellular-resolution connectomics aims to generate comprehensive brain connectivity maps; a key challenge is developing scalable algorithms; deep learning, specifically FusionNet, has shown promise in automatically segmenting neuronal structures in connectomics data.
FRONTIERS IN COMPUTER SCIENCE
(2021)
Article
Remote Sensing
Moulay A. Akhloufi et al.
Summary: Wildfires pose significant risks including economic losses, human casualties and environmental damage. Research has focused on developing dedicated solutions for wildland fire assistance and fighting, with proposed systems for remote detection and tracking showing improvements in data collection and fire characterization in small-scale environments. However, ground-based systems are limited in coverage for large wildfires, leading to the proposal of UAVs and UAS as a low-cost alternative for prevention, detection and real-time support of firefighting.
Article
Computer Science, Artificial Intelligence
Maria Joao Sousa et al.
EXPERT SYSTEMS WITH APPLICATIONS
(2020)
Article
Computer Science, Artificial Intelligence
Li Liu et al.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2020)
Review
Engineering, Multidisciplinary
Anshul Gaur et al.
Article
Computer Science, Artificial Intelligence
Xuebin Qin et al.
PATTERN RECOGNITION
(2020)
Article
Environmental Sciences
Panagiotis Barmpoutis et al.
Proceedings Paper
Computer Science, Information Systems
Haiyu Wu et al.
2020 54TH ANNUAL CONFERENCE ON INFORMATION SCIENCES AND SYSTEMS (CISS)
(2020)
Proceedings Paper
Computer Science, Hardware & Architecture
Byron Arteaga et al.
2020 IEEE INTERNATIONAL SYMPOSIUM ON SIGNAL PROCESSING AND INFORMATION TECHNOLOGY (ISSPIT 2020)
(2020)
Proceedings Paper
Remote Sensing
H. Harkat et al.
IMAGE AND SIGNAL PROCESSING FOR REMOTE SENSING XXVI
(2020)
Article
Environmental Sciences
Minsoo Park et al.
Review
Chemistry, Analytical
Panagiotis Barmpoutis et al.
Article
Computer Science, Artificial Intelligence
Songbin Li et al.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2020)
Article
Computer Science, Artificial Intelligence
Ziyang Tang et al.
Review
Engineering, Electrical & Electronic
Anshul Gaur et al.
IEEE SENSORS JOURNAL
(2019)
Review
Computer Science, Artificial Intelligence
Zhong-Qiu Zhao et al.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2019)
Proceedings Paper
Automation & Control Systems
Guangyi Wang et al.
PROCEEDINGS OF THE 2019 31ST CHINESE CONTROL AND DECISION CONFERENCE (CCDC 2019)
(2019)
Proceedings Paper
Engineering, Industrial
Yanhong Chen et al.
PROCEEDINGS OF THE 2019 14TH IEEE CONFERENCE ON INDUSTRIAL ELECTRONICS AND APPLICATIONS (ICIEA 2019)
(2019)
Proceedings Paper
Engineering, Electrical & Electronic
Dzemil Dzigal et al.
2019 11TH INTERNATIONAL CONFERENCE ON ELECTRICAL AND ELECTRONICS ENGINEERING (ELECO 2019)
(2019)
Proceedings Paper
Acoustics
Panagiotis Barmpoutis et al.
2019 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)
(2019)
Article
Computer Science, Interdisciplinary Applications
Wang Yuanbin et al.
JOURNAL OF ALGORITHMS & COMPUTATIONAL TECHNOLOGY
(2019)
Article
Chemistry, Analytical
Yi Zhao et al.
Proceedings Paper
Computer Science, Artificial Intelligence
Shixiao Wu et al.
2018 11TH INTERNATIONAL SYMPOSIUM ON COMPUTATIONAL INTELLIGENCE AND DESIGN (ISCID), VOL 1
(2018)
Proceedings Paper
Computer Science, Artificial Intelligence
Moulay A. Akhloufi et al.
PATTERN RECOGNITION AND TRACKING XXIX
(2018)
Article
Engineering, Civil
Tom Toulouse et al.
FIRE SAFETY JOURNAL
(2017)
Article
Computer Science, Artificial Intelligence
Vijay Badrinarayanan et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Gao Huang et al.
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017)
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Kaiming He et al.
2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Francois Chollet
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017)
(2017)
Article
Engineering, Electrical & Electronic
Tom Toulouse et al.
SIGNAL IMAGE AND VIDEO PROCESSING
(2016)
Article
Chemistry, Analytical
Henry Cruz et al.
Article
Engineering, Electrical & Electronic
Kosmas Dimitropoulos et al.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2015)
Article
Computer Science, Artificial Intelligence
Tom Toulouse et al.
IET IMAGE PROCESSING
(2015)
Article
Computer Science, Artificial Intelligence
Olga Russakovsky et al.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2015)
Proceedings Paper
Computer Science, Information Systems
Marcos Vinicius Naves Bedo et al.
ENTERPRISE INFORMATION SYSTEMS (ICEIS 2015)
(2015)
Proceedings Paper
Computer Science, Artificial Intelligence
Ross Girshick
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)
(2015)
Proceedings Paper
Computer Science, Artificial Intelligence
Daniel Y. T. Chino et al.
2015 28TH SIBGRAPI CONFERENCE ON GRAPHICS, PATTERNS AND IMAGES
(2015)
Article
Computer Science, Artificial Intelligence
Martin Mueller et al.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2013)
Article
Engineering, Civil
ByoungChul Ko et al.
FIRE SAFETY JOURNAL
(2010)
Article
Engineering, Multidisciplinary
L. Rossi et al.
MEASUREMENT SCIENCE AND TECHNOLOGY
(2010)
Article
Engineering, Civil
Turgay Celik et al.
FIRE SAFETY JOURNAL
(2009)
