检索结果-内蒙古大学图书馆

FireSense: Integrating deep learning with ESP32 Microcontrollers for Enhanced Forest Fire Surveillance 48

FireSense: Integrating Deep Learning with ESP32 Microcontrol...

48th Annual IEEE International Computers, Software, and Applications Conference (COMPSAC) - Digital Development for a Better Future

作者： Arunkumar, Rohan Dubey, Atul Dr Ronald E McNair Acad High Sch Jersey City NJ 07302 USA AIClub Res Inst Mountain View CA USA

ISBN: (纸本)9798350376975;9798350376968

The escalating impacts of climate change have notably increased the frequency and intensity of forest fires, presenting profound challenges to both environmental sustainability and socio-economic stability. Traditional wildfire management strategies, heavily reliant on suppression tactics, are increasingly inadequate due to their high resource demands and limited adaptability to rapid environmental changes. This study introduces an approach to wildfire detection by integrating advanced deep learning algorithms with practical hardware deployment. Utilizing the FLAME (Fire Luminosity Airborne-based Machine learning Evaluation) dataset, we develop a predictive model using lightweight deep learning architectures like MobileNetV2, EfficientNetB0, NASNetMobile, and ResNetV2101, which are optimized for real-time processing and low computational costs. We deployed this model on the cloud and we used an ESP32 microcontroller-based device equipped with a camera and servo motor, designed to autonomously monitor forest areas and provide immediate fire alerts. This integrated system not only aims to enhance early fire detection but also helps wildfire management to be more proactive and efficient. Our results demonstrate the potential of this technology to significantly reduce the time between fire onset and detection, thereby enabling quicker response times and reducing the overall impact of wildfires. Experimental results show that MobileNetV2 resulted in the best-performing model with a test accuracy of 98.1%, and the device performed up to expectations in hands-on testing scenarios. This research contributes significantly to the fields of environmental monitoring and disaster management, providing a scalable and effective solution to a critical global challenge intensified by climate change. Our work underscores the potential of this technology to reduce the time from fire detection to response, thereby mitigating the impact of wildfires.

关键词： deep learning transfer learning forest fire hardware device esp32 microcontroller FLAME dataset

来源：评论

学校读者我要写书评

暂无评论

Integration of Dynamic Slow Feature Analysis and deep Neural Networks for Subway Indoor PM2.5 Prediction

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2024年 73卷

作者： Lu, Yifeng Wang, Jinyong Liu, Tianlong Yoo, ChangKyoo Liu, Hongbin Nanjing Forestry Univ Jiangsu Coinnovat Ctr Efficient Proc & Utilizat Fo Nanjing 210037 Peoples R China Western Univ Dept Chem & Biochem Engn London ON N6A 3K7 Canada Kyung Hee Univ Coll Engn Dept Environm Sci & Engn Yongin 446701 South Korea

This study addresses the limitations in data-driven PM2.5 concentration prediction, which typically depends on statistical relationships with other factors, posting challenges in processing. To address the high costs associated with hardware-based monitoring, we introduced a novel hybrid model that synergizes dynamic slow feature analysis (DSFA), long short-term memory (LSTM) network, and convolutional block attention module (CBAM). The DSFA effectively resolves time lag issues prevalent in real industrial processes. When provided as input to the LSTM for training, the retrieved slow features effectively extract dynamic information from the data while minimizing complexity. Subsequently, CBAM adaptively adjusts feature weights, leading to refined prediction results. Comparative analysis reveals that our DSFA-LSTM-CBAM model outperforms conventional deep learning models, including partial least square (PLS), CNN, standard LSTM, and other hybrid models in predictive accuracy. Specifically, the model achieves a 45.6% reduction in root mean square error (RMSE) compared to the single LSTM model, and a 12.4% improvement in the coefficient of determination relative to the hybrid PCA-LSTM. In addition, this hybrid model demonstrates an enhanced capacity for handling nonlinearity and time-variability in time series data and exhibits strong robustness, marking a significant advancement in indoor air quality (IAQ) modeling.

关键词： Convolutional block attention module (CBAM) dynamic slow feature analysis (DSFA) indoor air quality (IAQ) long short-term memory (LSTM) PM2.(5) forecasting

来源：评论

学校读者我要写书评

暂无评论

Plant Disease Detection through External Camera Sensors using deep learning Techniques

Plant Disease Detection through External Camera Sensors usin...

引用

2024 International Conference on Electrical, Electronics and Computing Technologies, ICEECT 2024

作者： Thakur, Uttkarsh Srivastava, Prabhat Kumar Ghosh, Dipak Kumar Arora, Amit Galgotias University Dept. of Electrical Electronics & Communication Engineering Greater Noida India

ISBN: (纸本)9798350378092

Agriculture is one of the most important aspect of a country for its economy and livelihood. A country is dependent on its farmers for livelihood and for providing raw material for food. Therefore, crop yield is extremely essential. One of things that interfere with good crop yield is diseases in plants. Diseases in plants can occur due to variety of factors like temperature, fertility of soil, type of soil, pH of the soil and moisture of the soil. Quick and accurate detection is extremely necessary for eradication of plant diseases. Traditionally, disease of the plant is identified by visually studying the leaf by the naked eye. But this process takes considerable time of the farmer and may not be the efficient method for identifying diseases. image processing techniques, along with a classification algorithm solves this problem by observing patterns on plant leaves. We aim to build a solution which is able to capture image through external camera sources like mobile camera, webcam and extract features with image processing techniques and classify disease with deep Neural Network Algorithm. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Leaf Lens: An Intelligent Vision for Plant Disease Diagnosis using deep learning 2

Leaf Lens: An Intelligent Vision for Plant Disease Diagnosis...

引用

2nd International Conference on Artificial Intelligence and Machine learning Applications, AIMLA 2024

作者： Prasad, P. Yogendra Ramu, M. Reshma, Sanjamala Priya, Rayachoty Chandana Anusha, Pasupuleti Reddy, B Likhith Dept of CSSE Tirupati India Annamacharya Institute of Technology & Sciences Dept of CSE Tirupati India

ISBN: (纸本)9798350349221

In India, where 70% of the population is involved in agriculture, accurate recognition of botanical disorders is crucial to minimize crop losses. Manual monitoring of these diseases requires significant labor, expertise in plant pathology, and time investment. This paper presents a methodology for identifying plant diseases by analyzing leaves' images. image processing extracts valuable properties and information from the images, while deep learning, [3] a part of artificial intelligence, which simplifies data preprocessing. The algorithms used can process unstructured data, including text and images, automating feature extraction and diminishing the dependance on human professionals. This automated approach facilitates informed decision-making and precise predictions based on extensive training data. The paper introduces an innovative approach to plant disease detection through deep learning, specifically using the VGG19 model, which can scrutinize intricate patterns in plant images, providing a reliable method for identifying diseases. © 2024 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Integrating deep learning and Discrete Cosine Transform for DC-Resistivity Inversion: Application to Synthetic and real Data 6

Integrating Deep Learning and Discrete Cosine Transform for ...

引用

6th Asia Pacific Meeting on Near Surface Geoscience and Engineering: Smart Technologies Kind to the Planet

作者： Rincón, F. Aleardi, M. Stucchi, E. Tognarelli, A. Università di Pisa Italy

ISBN: (纸本)9789462824997

Electrical resistivity tomography (ERT) is a geophysical method crucial for subsurface imaging, offering key insights into fluid saturation and porosity. Large-scale ERT campaigns can pose logistical and time challenges related to data acquisition and processing. The ERT inverse problem is nonlinear and ill-posed, and it is usually solved through deterministic gradient-based algorithms. In this work, we develop a novel DC resistivity inversion approach that integrates a deep-learning framework with Discrete Cosine Transform (DCT). Furthermore, we incorporated a high-quality dataset featuring realistic resistivity models derived from real data. We first apply our approach to synthetic data not previously used during the training stage. Then, we demonstrate the applicability of the method through inverting real data. Our results show very good model prediction performance associated with low data misfit when applying the network to both synthetic and real data cases. Our inversion proposal provides an opportunity to obtain real-time resistivity models with minimal data misfit, thereby influencing decision-making during the acquisition stage and substantially reducing ERT-related costs for both acquisition and processing in large campaigns. © 2024 6th Asia Pacific Meeting on Near Surface Geoscience and Engineering: Smart Technologies Kind to the Planet.

关键词： Discrete cosine transforms

来源：评论

学校读者我要写书评

暂无评论

Sustainable Aquaculture Management through IoT and deep learning-Driven Remote Monitoring 8

Sustainable Aquaculture Management through IoT and Deep Lear...

引用

8th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2024

作者： Panduranga, D. Kumar, R.Prashanth Das, More Swami Narender, G. Sreelaxmi, M. Rajeswaran, N. Telangana Secunderabad India Marri Laxman Reddy Institute of Technology and Management Department of Cse Telangana Hyderabad India Cvr College of Engineering Department of Cse Telangana Hyderabad India Keshav Memorial Institute of Technology Department of It Telangana Hyderabad India Telangana Hyderabad India Malla Reddy College of Engg. Department of Ece Telangana Secunderabad India

ISBN: (纸本)9798350367904

Aquaculture, as a critical component of global food security, faces challenges in maintaining optimal conditions for aquatic organisms. The paper presents an innovative IoT and deep learning-based remote monitoring system designed to enhance sustainable aquaculture management. In order to monitor water quality factors including temperature, pH, dissolved oxygen, and feeding behavior in real time, the system incorporates a network of sensors. These sensors' data are locally processed on edge devices before being sent to a cloud-based platform for centralized storage and analysis. Convolutional neural networks (CNNs), in particular, are used in deep learning algorithms for image identification tasks to track the health and activity of fish. The system utilizes YOLOv8 for real-time detection and classification of fish diseases, ensuring early intervention and minimizing losses. User-friendly interfaces provide actionable insights through smart dashboards displaying trends, alerts, and recommended actions based on predictive analytics. By facilitating proactive management decisions, this IoT and deep learning-based system aims to optimize resource utilization, improve productivity, and promote sustainable practices in aquaculture. Case studies demonstrate its efficacy in enhancing operational efficiency, reducing environmental impact, and fostering long-term viability in aquatic food production systems. © 2024 IEEE.

关键词： Aquaculture

来源：评论

学校读者我要写书评

暂无评论

real-time location of acupuncture points based on anatomical landmarks and pose estimation models

引用

FRONTIERS IN NEUROROBOTICS 2024年 18卷 1484038页

作者： Malekroodi, Hadi Sedigh Seo, Seon-Deok Choi, Jinseong Na, Chang-Soo Lee, Byeong-il Yi, Myunggi Pukyong Natl Univ Ind 4-0 Convergence B Engn Busan South Korea Pukyong Natl Univ Div Smart Healthcare Major Biomed Engn Busan South Korea Dongshin Univ Coll Korean Med Naju South Korea Pukyong Natl Univ Div Smart Healthcare Major Human Bioconvergence Busan South Korea Pukyong Natl Univ Inst Informat Technol & Convergence Digital Healthcare Res Ctr Busan South Korea

Introduction Precise identification of acupuncture points (acupoints) is essential for effective treatment, but manual location by untrained individuals can often lack accuracy and consistency. This study proposes two approaches that use artificial intelligence (AI) specifically computer vision to automatically and accurately identify acupoints on the face and hand in real-time, enhancing both precision and accessibility in acupuncture *** The first approach applies a real-time landmark detection system to locate 38 specific acupoints on the face and hand by translating anatomical landmarks from image data into acupoint coordinates. The second approach uses a convolutional neural network (CNN) specifically optimized for pose estimation to detect five key acupoints on the arm and hand (LI11, LI10, TE5, TE3, LI4), drawing on constrained medical imaging data for training. To validate these methods, we compared the predicted acupoint locations with those annotated by *** Both approaches demonstrated high accuracy, with mean localization errors of less than 5 mm when compared to expert annotations. The landmark detection system successfully mapped multiple acupoints across the face and hand even in complex imaging scenarios. The data-driven approach accurately detected five arm and hand acupoints with a mean Average Precision (mAP) of 0.99 at OKS 50%.Discussion These AI-driven methods establish a solid foundation for the automated localization of acupoints, enhancing both self-guided and professional acupuncture practices. By enabling precise, real-time localization of acupoints, these technologies could improve the accuracy of treatments, facilitate self-training, and increase the accessibility of acupuncture. Future developments could expand these models to include additional acupoints and incorporate them into intuitive applications for broader use.

关键词： deep learning acupuncture traditional medicine computer vision pose estimation

来源：评论

学校读者我要写书评

暂无评论

ActivMedica—MRI Brain image Detection & Captioning with ActivGPT Integration

ActivMedica—MRI Brain Image Detection & Captioning with Act...

引用

International Conference on Advances and Applications of Artificial Intelligence and Machine learning, ICAAAIML 2023

作者： Malayil, Shanid Krishna, P. Sudeep Pattupara, Rabeen Mohamed, P. M. Rishil Faris, Salman Suhair, K. Muhammed Ahmed, P. Jemsheer MEA Engineering College Kerala Perinthelmanna India

ISBN: (纸本)9789819795062

ActivMedica is an innovator in brain tumor diagnosis in the rapidly changing healthcare industry, employing cutting edge AI tools including deep learning and natural language processing (NLP). The shortcomings of existing manual approaches for medical imaging data analysis include difficulty in identifying diverse brain tumors and increasing medical expenditures. Our approach, which at first is based on transfer learning and GIT architecture, uses the GIT convolutional neural network, or CNN, model. A carefully chosen and preprocessed diverse dataset, enhanced for variety and improved by transfer learning from the imageNet dataset for feature extraction, is essential for successful model training. Using the GIT-based methodology, ActivGPT, an AI chatbot built on OpenAI, enables real-time image analysis through natural language conversations. In addition to brain tumor detection, our research generates brief medical reports based on detected brain tumors, providing medical experts with comprehensive information. Brain tumor diagnosis is ethically improved by ActivMedica, a comprehensive healthcare innovation that combines AI, deep learning, and natural language processing. By doing this, diagnosis accuracy is greatly increased while moral and ethical requirements for AI use in healthcare are maintained. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

real-time tracking of fast-moving object in occlusion scene

引用

Journal of Systems Engineering and Electronics 2024年第3期35卷 741-752页

作者： LI Yuran LI Yichen ZHANG Monan YU Wenbin GUAN Xinping Department of Automation Shanghai Jiao Tong UniversityShanghai 200240China Key Laboratory of Systems Control and Information Processing Ministry of Education of ChinaShanghai 200240China

Tracking the fast-moving object in occlusion situations is an important research topic in computer vision. Despite numerous notable contributions have been made in this field,few of them simultaneously incorporate both object's extrinsic features and intrinsic motion patterns into their methodologies,thereby restricting the potential for tracking accuracy improvement. In this paper, on the basis of efficient convolution operators(ECO) model, a speed-accuracy-balanced model is put forward. This model uses the simple correlation filter to track the object in real-time, and adopts the sophisticated deep-learning neural network to extract high-level features to train a more complex filter correcting the tracking mistakes, when the tracking state is judged to be poor. Furthermore, in the context of scenarios involving regular fast-moving, a motion model based on Kalman filter is designed which greatly promotes the tracking stability, because this motion model could predict the object's future location from its previous movement pattern. Additionally,instead of periodically updating our tracking model and training samples, a constrained condition for updating is proposed,which effectively mitigates contamination to the tracker from the background and undesirable samples avoiding model degradation when occlusion happens. From comprehensive experiments, our tracking model obtains better performance than ECO on object tracking benchmark 2015(OTB100), and improves the area under curve(AUC) by about 8% and 32% compared with ECO, in the scenarios of fast-moving and occlusion on our own collected dataset.

关键词： speed-accuracy balanced motion modeling constrained updater

来源：评论

学校读者我要写书评

暂无评论

ACCURATE NON-DESTRUCTIVE TESTING METHOD FOR POTATO SPROUTS FOCUSING ON DEFORMABLE ATTENTION

引用

INMATEH-AGRICULTURAL ENGINEERING 2024年第1期72卷 402-413页

作者： Geng, Binxuan Dai, Guowei Zhang, Huan Qi, Shengchun Dewi, Christine Qingdao Agr Univ Fac Mech & Elect Engn Qingdao 266000 Peoples R China Agr Informat Inst CAAS Beijing 100081 Peoples R China Satya Wacana Christian Univ Fac Informat Technol Salatiga Indonesia

Accurate potato sprout detection is the key to automatic seed potato cutting, which is important for potato quality and yield. In this paper, a lightweight DAS-YOLOv8 model is proposed for the potato sprout detection task. By embedding DAS deformable attention in the feature extraction network and the feature fusion network, the global feature context can be efficiently represented and the attention increased to the relevant pixel image region;then, the C2f_Atten module fusing Shuffle attention is designed based on the C2f module to satisfy the attention to the key feature information of the high-level abstract semantics of the feature extraction network. At the same time, the ghost convolution is introduced to improve the C2f module and convolutional module to realize the decomposition of the redundant features to extract the key features. Verified on the collected potato sprout image data set, the average accuracy of the proposed DAS-YOLOv8 model is 94.25%, and the calculation amount is only 7.66 G. Compared with the YOLOv8n model, the accuracy is 2.13% higher, and the average accuracy is 1.55% higher. In comparison to advanced state-of-the-art (SOTA) target detection algorithms, the method in this paper offers a better balance between comprehensive performance and lightweight model design. The improved and optimized DAS-YOLOv8 model can realize the effective detection of potato sprouts, meet the requirements of real-time processing, and can provide theoretical support for the non-destructive detection of sprouts in automatic seed potato cutting.

关键词： deep learning non-destructive inspection YOLOv8 DAS Attention Mechanism potato sprouts

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：