Remote sensing scene classification has been extensively studied for its critical roles in geological survey, oil exploration, traffic management, earthquake prediction, wildfire monitoring, and intelligence monitorin...
详细信息
ISBN:
(纸本)9781510666931;9781510666948
Remote sensing scene classification has been extensively studied for its critical roles in geological survey, oil exploration, traffic management, earthquake prediction, wildfire monitoring, and intelligence monitoring. In the past, the machine Learning (ML) methods for performing the task mainly used the backbones pretrained in the manner of supervised learning (SL). As Masked image Modeling (MIM), a self-supervised learning (SSL) technique, has been shown as a better way for learning visual feature representation, it presents a new opportunity for improving ML performance on the scene classification task. This research aims to explore the potential of MIM pretrained backbones on four well-known classification datasets: Merced, AID, NWPU-RESISC45, and Optimal-31. Compared to the published benchmarks, we show that the MIM pretrained vision Transformer (viTs) backbones outperform other alternatives (up to 18% on top 1 accuracy) and that the MIM technique can learn better feature representation than the supervised learning counterparts (up to 5% on top 1 accuracy). Moreover, we show that the general-purpose MIM-pretrained viTs can achieve competitive performance as the specially designed yet complicated Transformer for Remote Sensing (TRS) framework. Our experiment results also provide a performance baseline for future studies.
Weeds are unwanted plants that grow with crops and usually removed by spraying herbicides or by manual labour. Herbicides being sprayed mostly do not reach their target because of the focus on a very wide area. This a...
详细信息
Weeds are unwanted plants that grow with crops and usually removed by spraying herbicides or by manual labour. Herbicides being sprayed mostly do not reach their target because of the focus on a very wide area. This also tends to harm the environment, and other living organisms. Manual labour is time-consuming and expensive and it is continuously managed and monitored. The autonomous robotics and imageprocessing tasks can be completed with precision and ease in agriculture. With imageprocessing, plants and weeds can be classified. Methods like scale invariant feature transforms (SIFT), speeded-up robust features (SURF), and ensemble learning, neural networks can be incorporated into identifying the difference. We can easily classify weeds and crops from images of plantations leveraging machine learning algorithms, artificial vision analysis systems, among others. Deep learning methods like convolutional neural network (CNN), rectified linear units (ReLU) and SoftMax (for classification) are focused in this paper.
The proceedings contain 128 papers. The special focus in this conference is on Data Science, machine Learning and applications. The topics include: Digitization of Monuments – An Impact on the Tourist Experience with...
ISBN:
(纸本)9789819780303
The proceedings contain 128 papers. The special focus in this conference is on Data Science, machine Learning and applications. The topics include: Digitization of Monuments – An Impact on the Tourist Experience with Special Reference to Hampi;resume Parser Using machine Learning;IOT Based Smart Hydroponics System;comparative Study of machine Learning and Deep Learning Techniques for Cancer Disease Detection;High Thruput Modulation Approaches Used in Next Generation WiF’s Under Multi-impairments Environments with MATLAB Codes;skin Disease Detection;root vegetable Crop Recommendation System Based on Soil Properties and Environmental Factors;deep Learning Model Development for an Automatic Healthcare Edge Computing Application;Empathetic Conversations in Mental Health: Fine-Tuning LLMs for Supportive AI Interactions;exploring Block Chain Technology with applications, and Future Prospects;a Comprehensive Review of Soft Computing Enabled Techniques for IoT Security: State-of-the-Art and Challenges Ahead;Performance Analysis of machine Learning Algorithms on Imbalanced Datasets Using SMOTE Technique;An AI Based Nutrient Tracking and Analysis System;power Saving Mechanism for Street Lights System Using IoT;Automatic Login System Using ATTINY85 IC;forecasting Stock Prices: A Comparative Analysis of machine Learning, Deep Learning, and Statistical Approaches;smart vision Bot;robots in Logistics: Apprehension of Current Status and Future Trends in Indian Warehouses;smart Healthcare: Enhancing Patient Well-Being with IoT;Detection of B-ALL Using CNN Model and Deep Learning;a Comprehensive Analysis for Advancements and Challenges in Deep Learning Models for imageprocessing;a Comprehensive Survey on Enhancing Patient Care Through Deep Learning and IoT-Enabled Healthcare Innovations;attention-Based image Caption Generation.
The omnipresence and deep impact of artificial intelligence (AI) in today's society are undeniable. While the technology has already established itself as a powerful tool in several industries, more recently it ha...
详细信息
The omnipresence and deep impact of artificial intelligence (AI) in today's society are undeniable. While the technology has already established itself as a powerful tool in several industries, more recently it has also started to change the practice of medicine. The aim of this review is to provide healthcare providers working in the field of cardiovascular medicine with an overview of AI and machine learning (ML) algorithms that have passed the initial tests and made it into contemporary clinical practice. The following domains where AI/ML could revolutionize cardiology are covered: (i) signal processing, (ii) imageprocessing, (iii) clinical risk stratification, (iv) natural language processing, and (v) fundamental clinical discoveries.
Transformer-based Deep Neural Network architectures have gained tremendous interest due to their effectiveness in various applications across Natural Language processing (NLP) and Computer vision (Cv) domains. These m...
详细信息
Transformer-based Deep Neural Network architectures have gained tremendous interest due to their effectiveness in various applications across Natural Language processing (NLP) and Computer vision (Cv) domains. These models are the de facto choice in several language tasks, such as Sentiment Analysis and Text Summarization, replacing Long Short Term Memory (LSTM) model. vision Transformers (viTs) have shown better model performance than traditional Convolutional Neural Networks (CNNs) in visionapplications while requiring significantly fewer parameters and training time. The design pipeline of a neural architecture for a given task and dataset is extremely challenging as it requires expertise in several interdisciplinary areas such as signal processing, imageprocessing, optimization and allied fields. Neural Architecture Search (NAS) is a promising technique to automate the architectural design process of a Neural Network in a data-driven way using machine Learning (ML) methods. The search method explores several architectures without requiring significant human effort, and the searched models outperform the manually built networks. In this paper, we review Neural Architecture Search techniques, targeting the Transformer model and its family of architectures such as Bidirectional Encoder Representations from Transformers (BERT) and vision Transformers. We provide an in-depth literature review of approximately 50 state-of-the-art Neural Architecture Search methods and explore future directions in this fast-evolving class of problems.
The exploration of sentiments through facial expressions is a captivating domain with applications across security, healthcare, and human–computer interaction, where understanding sentiments is primarily about interp...
详细信息
The proceedings contain 173 papers. The topics discussed include: restricted area sign detector using YOLO v5;research on distance teaching course interactive system based on computer algorithm research data;APT detec...
ISBN:
(纸本)9798350323313
The proceedings contain 173 papers. The topics discussed include: restricted area sign detector using YOLO v5;research on distance teaching course interactive system based on computer algorithm research data;APT detection and attack scenario reconstruction based on big data analysis;new imageprocessing: vGG image style transfer with gram matrix style features;trajectory measurement and positioning of underwater vehicle based on monocular stereo vision;the importance of multi feature extraction and fusion for prediction of protein subcellular localization;design and implementation of FPGA-based four-dimensional ultra chaotic system;flocking towards a robust mobile network topology;real time speech recognition method for online complaints from power grid customers based on improved residual network;optimization of parking space detection system based on ZigBee wireless sensor network;and a wire drawing defect detection approach for FDM 3D printing based on machinevision technology.
Deep learning algorithms have shown exceptional effectiveness in a wide range of supervised and unsupervised learning tasks in a variety of fields, including imageprocessing, computer vision, natural language process...
详细信息
Deep learning algorithms have shown exceptional effectiveness in a wide range of supervised and unsupervised learning tasks in a variety of fields, including imageprocessing, computer vision, natural language processing, and speech or voice processing. In this paper, a comprehensive analysis is conducted to assess the impact of deep learning on user authentication using both physiological and behavioural biometrics. This work encompasses the diverse deep learning approaches employed in authentication schemes tailored for smart devices. Meticulous scrutiny of commonly used datasets in these authentication studies is undertaken, accompanied by a comparative assessment of performance. The deep learning models under consideration span a spectrum of architectures, including deep neural networks, convolutional neural networks, deep auto-encoders, recurrent neural networks, and their variants. To enhance the clarity and categorization of authentication techniques for smart devices, a taxonomy is proposed based on the specific authentication metrics employed: (1) Knowledge-based Authentication (KBA), (2) Physiological Biometric-based Authentication (PBBA), (3) Behavioural Biometric-based Authentication (BBBA), (4) Physiological and Behavioural Continuous Authentication (PBBCA), and (5) Multi Modal Authentication (MMA). Furthermore, potential contributions of deep learning techniques to the realms of physiological and behavioural biometrics are discussed. Significance is placed on performance metrics, including accuracy, stability, and robustness, in evaluating these deep learning-based authentication systems. The challenges and limitations that deep learning approaches must surmount when dealing with real-world biometric data in the context of biometric identification systems are examined. This work not only underscores the transformative role of deep learning in user authentication but also offers valuable insights into the evolving landscape of biometric identification o
This paper aims to design and implement an MLbased approach to learn from NeuroAqua - the AI and IoT-based aquaponics system set up in our previous research at both a lab setting and larger-scale Ouroboros Aquaponics ...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
This paper aims to design and implement an MLbased approach to learn from NeuroAqua - the AI and IoT-based aquaponics system set up in our previous research at both a lab setting and larger-scale Ouroboros Aquaponics Farm (Half Moon Bay, CA) to enhance system stability and efficiency. Utilizing the data gathered from the wireless sensors, a structured database was formed to store the aquaponics environmental conditions, water quality, nutrient components, and plant images. We used the ML model to find the important factors having the largest impact on plant growth and their optimal amount levels. First, computer vision with imageprocessing was applied to develop auto plant growth monitoring and to measure plant growth rate as the target variable more accurately and automatically for ML. Then feature engineering on the input variables was performed to enhance model performance and accuracy for a smaller dataset. ML algorithms including Linear Regression, Bagging Regressor, Decision Tree, Random Forest, XGBoost and Artificial Neural Network were applied and evaluated based on key performance metrics. The findings show that XGBoost outperformed the other models with 91.6% accuracy and also had the lowest MAE. Random Forest came in second with 90.9% accuracy and then Bagging Regressor in third with 88.5% accuracy. Lastly, according to the feature importance analysis conducted on the best model of XGBoost, Nitrogen had the largest impact on plant growth, followed by Nitrate, Nitrite, Light, and Phosphorus. Hence the initial results would recommend to closely monitor these top important factors together with plant growth in NeuroAqua's monitoring applications.
In the course of modernization of camera based imaging and image analysis for accelerator hardware and beam control at the ELSA facility, a distributed imageprocessing approach was implemented, called FGrabbit. We ut...
详细信息
暂无评论