检索结果-内蒙古大学图书馆

image-Based Air Quality Estimation Using Convolutional neural Network Optimized by Genetic Algorithms: A Multi-Dataset Approach

引用

INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND applications 2025年第3期16卷 1177-1185页

作者： Khan, Arshad Ali Majid, Mazlina Abdul Dandoush, Abdulhalim Univ Malaysia Pahang Al Sultan Abdullah Fac Comp Pekan Malaysia Ctr Artificial Intelligence & Data Sci Gambang Malaysia Univ Doha Sci & Technol UDST Doha Qatar

Air pollution poses significant threats to human health and the environment, making effective monitoring increasingly essential. Traditional methods using fixed monitoring stations have challenges related to high costs and limited coverage. This paper proposes a new approach using convolutional neural networks with genetic algorithms for estimating air quality directly from images. The convolutional neural network is optimized using genetic algorithms, which dynamically tune hyper-parameters such as learning rate, batch size, and momentum to improve performance and generalizability across diverse environmental conditions. Our approach improves performance and reduces the risk of overfitting, thus ensuring balanced and robust results. To mitigate overfitting, we implemented dropout layers, batch normalization, and early stopping, significantly enhancing the model's generalization capability. Specifically, three different open-access datasets were combined into a single training dataset, capturing extensive temporal, spatial, and environmental variability. Extensive testing of the model performance was conducted with a broad set of metrics, including precision, recall, and F1 score. The results demonstrate that our model not only achieves high accuracy but also maintains well-balanced performance across all metrics, ensuring robust classification of different air quality levels. For instance, the model achieved a precision of 0.97, a recall of 0.97, and an overall accuracy of 0.9544 percent, outperforming baseline methods significantly in all metrics. These improvements underscore the effectiveness of Genetic Algorithms in optimizing the model.

关键词： Convolutional neural network Genetic Algorithm air quality estimation image processing

来源：评论

学校读者我要写书评

暂无评论

Fpga-based SoC design for real-time facial point detection using deep convolutional neural networks with dynamic partial reconfiguration

引用

SIGNAL image AND VIDEO processing 2024年第SUPPL 1期18卷 599-615页

作者： Teboulbi, Safa Messaoud, Seifeddine Hajjaji, Mohamed Ali Mtibaa, Abdellatif Atri, Mohamed Univ Monastir Higher Inst Comp Sci & Math Monastir Tunisia Univ Monastir Fac Sci Monastir Monastir Tunisia Univ Sousse Higher Inst Appl Sci & Technol Sousse Sousse Tunisia Univ Sfax Natl Engn Sch Sfax Syst Integrat & Emerging Energies Lab LR21ES14 Sfax Tunisia King Khalid Univ Coll Comp Sci Abha 61421 Saudi Arabia

Deep convolutional neural networks (DCNNs) have been mainly powerful and important artificial intelligence techniques, which are exploited in various computer vision applications, such as facial point detection (FPD), owing their versatility and high performance. The role of DCNNs in this area is pivotal due to their ability to learn hierarchical features, which are essential for recognizing complex patterns. On the other hand, DCNNs have considerable computational complexity due to their topology nature. An FPGA-based SoC design is needed for CNN acceleration due to the rapid development, energy efficiency, low latency and high reconfigurability of FPGAs. Despite their performance benefits, FPGA-based SoCs also come with several limitations like complexity of design. This paper proposes a dynamic partial reconfiguration (DPR) and hybrid architecture for DCNN accelerators. We propose (i) a GPU-based software implementation for DCNN-based FPD, (ii) a CNN-based acceleration and optimization method using the high-level synthesis technique, (iii) a DPR-based hybrid architecture to improve the performance of the suggested approach. To validate our design, four scenarios are put forward. The experimental results prove that the proposed work achieves a better performance in terms of reconfigurability and execution time, hardware cost, and power consumption.

关键词： CNN FPGA-SoC Acceleration Dynamic reconfiguration Co-design

来源：评论

学校读者我要写书评

暂无评论

Compression of Deep neural networks based on quantized tensor decomposition to implement on reconfigurable hardware platforms

引用

neural networks 2022年第0期150卷 350-363页

作者： Nekooei, Amirreza Safari, Saeed Univ Tehran Tehran Iran

Deep neural networks (DNNs) have been vastly and successfully employed in various artificial intelligence and machine learning applications (e.g., image processing and natural language processing). As DNNs become deeper and enclose more filters per layer, they incur high computational costs and large memory consumption to preserve their large number of parameters. Moreover, present processing platforms (e.g., CPU, GPU, and FPGA) have not enough internal memory, and hence external memory storage is needed. Hence deploying DNNs on mobile applications is difficult, considering the limited storage space, computation power, energy supply, and real-time processing requirements. In this work, using a method based on tensor decomposition, network parameters were compressed, thereby reducing access to external memory. This compression method decomposes the network layers' weight tensor into a limited number of principal vectors such that (i) almost all the initial parameters can be retrieved, (ii) the network structure did not change, and (iii) the network quality after reproducing the parameters was almost similar to the original network in terms of detection accuracy. To optimize the realization of this method on FPGA, the tensor decomposition algorithm was modified while its convergence was not affected, and the reproduction of network parameters on FPGA was straightforward. The proposed algorithm reduced the parameters of ResNet50, VGG16, and VGG19 networks trained with Cifar10 and Cifar100 by almost 10 times. (C)& nbsp;2022 Elsevier Ltd. All rights reserved.

关键词： Deep neural Network Tensor decomposition Principal vectors Network weights External memory

来源：评论

学校读者我要写书评

暂无评论

Applying Deep neural networks and NLP Techniques for Sentiment Analysis in Social Media Data 2

Applying Deep Neural Networks and NLP Techniques for Sentime...

引用

2nd International Conference on artificial Intelligence and Machine Learning applications, AIMLA 2024

作者： Brinda, B.M. Rajan, C. Geetha, K. Nathiya, N. Raguraman, P.J. Srinivasan, K. Paavai College of Engineering Computer Science and Engineering Namakkal India Information Technology KSRCT College of Technology Namakkal India Excel Engineering College Computer Science and Engineering Namakkal India Paavai College of Engineering Artificial Intelligence and Data Science Namakkal India

ISBN: (纸本)9798350349221

This work offers a comprehensive investigation of sentiment analysis in social media communication through the integration of deep learning techniques with a natural language processing (NLP) methodology. The goal of the project is to create a matching model that can be used in real-world social processes. This model will allow for the precise identification of pertinent content that is dynamically changing and the realtime selection of key phrases based on available data. The paper builds a message sentiment analysis model and an image message multimodal sentiment analysis model, exploring unimodal and multimodal sentiment analysis algorithms in social networks. The study shows how well it works to extract complex emotions from social media writing by fusing advanced deep neural networks - like transformers or recurrent neural networks - with natural language processing techniques. Furthermore, the study presents noteworthy results, such as 96.1% order accuracy for brief texts in deep learning models that have been optimized, sentiment filtering for positive, neutral, and negative comments in social network data that has been successful, and an assessment of sophisticated semantic similarity models that offers a thorough comprehension of their performance in classification tasks. This work provides insightful information that may be used to improve sentiment analysis models and their usefulness in tasks involving semantic similarity and social network data interpretation. © 2024 IEEE.

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Fusing Depths: Investigating the Synergy of Convolutional neural networks and Long Short-Term Memory networks for Enhanced image Caption Generation 1

Fusing Depths: Investigating the Synergy of Convolutional Ne...

引用

1st IEEE International Conference on Cognitive Robotics and Intelligent Systems, ICC - ROBINS 2024

作者： Sharma, Vikas Alekh Chaudhary, Kirti Vashishth, Tarun Kumar Chaudhary, Sachin Kumar, Bhupendra Iimt University School of Computer Science and Applications U.P Meerut India Delhi Ncr Ghaziabad India Gl Bajaj Institute of Management Greater Noida India

ISBN: (纸本)9798350372748

Generation, poses a challenge due to the intricate visual content and nuanced semantic details. This research introduces a novel approach for image captioning by seamlessly integrating Convolutional neural networks (CNNs) and Long Short-Term Memory (LSTM) networks. The proposed methodology involves leveraging CNNs to extract visual features and employing LSTM for the generation of descriptive captions. To further enhance the model's performance, attention mechanisms are incorporated, allowing the model to focus on relevant visual features during the caption generation process. The evaluation of our modified model utilizes standard benchmark datasets, such as Flickr8k, and employs metrics including CIDEr, METEOR, and BLEU scores. Through rigorous assessment, our model showcases improved performance, demonstrating its efficacy in comparison to existing methods. The versatility of the proposed system extends its potential applications to image retrieval, image description, and other multimedia scenarios requiring robust image analysis and natural language processing capabilities. This research contributes to the advancement of image captioning techniques, offering a promising solution for real-world applications in multimedia and artificial intelligence domains. The goal of this study is to explore and leverage the combined capabilities of CNNs and LSTMs to enhance the process of generating descriptive captions for images. By merging the strengths of CNNs in image feature extraction with the sequential understanding and context modeling abilities of LSTMs, the aim is to develop a more sophisticated and effective approach for generating accurate and contextually relevant captions that better capture the nuances and details of the images. This research seeks to push the boundaries of image captioning technology, ultimately improving the quality and richness of generated captions, and advancing the state-of-the-art in artificial intelligence and computer vision applicatio

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Implementation of Parallel Evolutionary Convolutional neural Network for Classification in Human Activity and image Recognition 22nd

Implementation of Parallel Evolutionary Convolutional Neural...

引用

22nd Mexican International Conference on artificial Intelligence (MICAI)

作者： Villegas-Cortez, Juan Roman-Alonso, Graciela Fernandez De Vega, Francisco Flores-Morales, Yafte Aaron Cordero-Sanchez, Salomon Univ Autonoma Metropolitana Dept Sistemas Azapotzalco Av Sn Pablo 180 Mexico City 02200 DF Mexico Univ Autonoma Metropolitana Dept Ingn Elect Unidad Iztapalapa Mexico City DF Mexico Univ Extremadura Dept Comp Sci C Santa Teresa Jornet38 Merida 06800 Spain Univ Autonoma Metropolitana Unidad Iztapalapa Dept Chem Mexico City DF Mexico

ISBN: (纸本)9783031477645;9783031477652

Pattern recognition has been evolving to include problems posed by new sceneries containing a high number of pattern components. processing this volume of information allows a more exact classification in wider types of applications;however, some of the difficulties of this scheme is the maintenance of numerical precision and mainly the reduction of the execution time. During the last 15 years, several Machine Learning solutions have been implemented to reduce the number of pattern components to be analyzed, such as artificial neural networks. Deep learning is an appropriate tool to accomplish this task. In this paper, a convolutional neural network is implemented for recognition and classification of human activity signals and digital images. It is achieved by automatically adjusting the parameters of the neural network through genetic algorithms using a multiprocessor and GPU platform. The results obtained show the reduction of computational costs and the possibility of better understanding of the solutions provided by Deep Learning.

关键词： Deep Learning Pattern recognition HAR image Recognition Parallel Genetic Algorithms

来源：评论

学校读者我要写书评

暂无评论

Dual-gate manipulation of a HfZrOx-based MoS2 field-effect transistor towards enhanced neural network applications

引用

NANOSCALE 2022年第1期15卷 313-320页

作者： Liu, Yilun Li, Qingxuan Zhu, Hao Ji, Li Sun, Qingqing Zhang, David Wei Chen, Lin Fudan Univ Sch Microelect State Key Lab AS & Syst Shanghai 200433 Peoples R China

artificial neural networks (ANNs) have strong learning and computing capabilities, and alleviate the problem of high power consumption of traditional von Neumann architectures, providing a solid basis for advanced image recognition, information processing, and low-power detection. Recently, a two-dimensional (2D) MoS2 field-effect transistor (FET) integrating a Zr-doped HfO2 (HZO) ferroelectric layer has shown potential for both logic and memory applications with low power consumption, which is promising for parallel processing of massive data. However, the long-term potentiation (LTP) characteristics of such devices are usually non-linear, which will affect the replacement of ANN weight values and degrade the ANN recognition rate. Here, we propose a dual-gate-controlled 2D MoS2 FET employing HZO gate stack with a crested symmetric structure to reduce power consumption. Improved nonlinearity of the LTP properties has been achieved through the electrical control of the dual gates. A recognition rate reaching 100% is obtained after 60 training epochs, and is 7.89% higher than that obtained from single-gate devices. Our proposed device structure and experimental results provide an attractive pathway towards high-efficiency data processing and image classification in the advanced artificial intelligence field.

关键词： Field effect transistors

来源：评论

学校读者我要写书评

暂无评论

Self-organising neural Discrete Representation Learning a la Kohonen 33rd

Self-organising Neural Discrete Representation Learning a la...

引用

33rd International Conference on artificial neural networks and Machine Learning (ICANN)

作者： Irie, Kazuki Csordas, Robert Schmidhuber, Juergen Harvard Univ Ctr Brain Sci Cambridge MA 02138 USA Stanford Univ Stanford CA 94305 USA USI Swiss AI Lab IDSIA Lugano Switzerland SUPSI Lugano Switzerland King Abdullah Univ Sci & Technol KAUST AI Initiat Thuwal Saudi Arabia IDSIA Lugano Switzerland

ISBN: (纸本)9783031723315;9783031723322

Unsupervised learning of discrete representations in neural networks (NNs) from continuous ones is essential for many modern applications. Vector Quantisation (VQ) has become popular for this, in particular in the context of generative models, such as Variational Auto-Encoders (VAEs), where the exponential moving average-based VQ (EMA-VQ) algorithm is often used. Here, we study an alternative VQ algorithm based on Kohonen's learning rule for the Self-Organising Map (KSOM;1982). EMA-VQ is a special case of KSOM. KSOM is known to offer two potential benefits: empirically, it converges faster than EMA-VQ, and KSOM-generated discrete representations form a topological structure on the grid whose nodes are the discrete symbols, resulting in an artificial version of the brain's topographic map. We revisit these properties by using KSOM in VQ-VAEs for image processing. In our experiments, the speed-up compared to well-configured EMA-VQ is only observable at the beginning of training, but KSOM is generally much more robust, e.g., w.r.t. the choice of initialisation schemes (Our code is public: https://***/IDSIA/kohonen-vae. The full version with an appendix can be found at: https://***/abs/2302.07950).

关键词： self-organizing maps Kohonen maps vector quantisation VQ-VAE discrete representation learning

来源：评论

学校读者我要写书评

暂无评论

Augmented GAN: Advancing Generative Adversarial networks Through Enhanced Discriminator Training and Generator Stability

Augmented GAN: Advancing Generative Adversarial Networks Thr...

引用

2024 International Conference on Signal processing, Computation, Electronics, Power and Telecommunication, IConSCEPT 2024

作者： Prakash, Aman Varghese, Ryan Binu Tanwar, Lavi Delhi Technological University Department of Electronics and Communications Engineering India

ISBN: (纸本)9798331540685

General Adversarial networks (GANs) have emerged as a powerful framework for generating reliable and transformative synthetic data in areas such as image generation, image and text synthesis, and data augmentation. This paper presents a comprehensive guide to building a Generative Adversarial neural Network using TensorFlow and Python. Through our research on this topic, we delve into the implementation details, providing step-by-step instructions for constructing both the generator and discriminator networks using TensorFlow, a popular deep-learning framework. Furthermore, we explore techniques for optimizing GAN performance, including architectural modifications, loss function selection, and training strategies. We demonstrate how to train a GAN on real-world datasets through practical examples, showcasing its capability to generate high-quality synthetic samples that closely resemble the training data distribution. In this paper, we aim to match fashion products such as clothing and accessories using GAN. This paper also focuses on improving the robustness and generalization of Convolutional neural Network (CNN) models, such as image classification and product recommendation. By combining image results with GANs and incorporating these artificial data into the training process, we aim to improve the performance of CNN models in real-world applications. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

CNNX: A Low Cost, CNN Accelerator for Embedded System in Vision at Edge

引用

ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING 2023年第2期48卷 1537-1545页

作者： Farahani, Ali Beithollahi, Hakem Fathi, Mahmood Barangi, Reza Iran Univ Sci & Technol Sch Comp Engn Tehran Iran

Convolutional neural networks (CNNs) have been widely deployed in artificial intelligence, including computer vision and pattern recognition. In these applications, CNN is the most computationally intensive part. Recently, many researchers have used depthwise convolution to decrease the computational load in the execution of CNNs;on the other hand, today, CNNs have become larger and larger. Consequently, they need more computational budget for their executions. The problem is more serious when this application is run in an embedded system, especially in the edge devices, as the embedded processor can hardly handle these heavy computational loads. This paper proposes a lightweight, low-power, and efficient CNN hardware accelerator for edge computing devices. This accelerator is explicitly designed for depthwise CNN. The proposed accelerator can be configured and programmed to run any lightweight CNN of a wide range of AI networks such as MobileNet, Xception, and shuffleNet. Our experimental results show that our accelerator can run MobileNet 70 times per second in a remote sensing AI application with a 224 x 224 pixel image from the imageNet dataset.

关键词： Convolutional neural networks (CNNs) Hardware accelerator Low power AI at edge Vision at edge Edge processing

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：