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, Hardware & Architecture
Irfan Younas et al.
Summary: This paper studies the sensor selection problem in IoT systems, treating it as a multi-objective optimization problem with 5 objectives. A Decomposition-based Many-Objective Evolutionary Algorithm and Opposition Based Learning are incorporated to solve the problem. Experimental results show that the proposed algorithm outperforms other compared algorithms.
COMPUTERS & ELECTRICAL ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Hong-Gui Han et al.
Summary: An accelerated second-order learning (ASOL) algorithm is proposed to train RBFNN, which reduces vanishing gradient, simplifies structure, and improves generalization ability through adaptive expansion and pruning mechanism. The theoretical analysis and experimental results demonstrate that ASOL-SORBFNN performs well in terms of learning speed and prediction accuracy.
Article
Computer Science, Artificial Intelligence
Lorraine Chambers et al.
Summary: Convolutional Neural Networks (CNNs) can achieve state of the art results for visual recognition problems when the data distributions are the same between train and test sets and all test set classes are present in the training data. However, in the real world, where data evolves and new classes emerge, traditional neural networks fail to identify unknown classes. Open-Set Classification research field provides potential solutions for this problem. In this study, a system called DeepStreamOS is proposed, which combines deep neural network activations with a stream-based outlier detection method to quickly identify instances belonging to unknown classes. Experimental results demonstrate that DeepStreamOS outperforms other open-set classification methods in most scenarios and significantly improves the speed of classification.
PATTERN RECOGNITION LETTERS
(2022)
Review
Chemistry, Analytical
Shibo Zhang et al.
Summary: The development of mobile and wearable devices has enabled various applications that measure and improve our daily lives, such as activity tracking, wellness monitoring, and human-computer interaction. Many of these applications rely on low-power sensors in these devices to perform human activity recognition (HAR). In recent years, deep learning has made significant advancements in HAR on mobile and wearable devices. This paper categorizes and summarizes existing work on deep learning methods for wearables-based HAR, and provides an analysis of the current advancements, developing trends, and major challenges. The paper also presents cutting-edge frontiers and future directions for deep learning-based HAR.
Article
Environmental Sciences
Zhen Zhang et al.
Summary: This article proposes a new paradigm for automatically designing a suitable CNN architecture for remote sensing scene classification. The more efficient RS-DARTS search framework is adopted to find the optimal network architecture, with new strategies introduced in the search phase, noise added to suppress skip connections, and sampling to reduce redundancy in exploring the network space. Extensive experiments demonstrate the effectiveness of the proposed method in classification performance and search cost compared to other methods.
Article
Computer Science, Information Systems
Nafiul Rashid et al.
Summary: This research introduces an adaptive convolutional neural network for energy-efficient human activity recognition on low-power edge devices. The proposed method outperforms traditional approaches in terms of energy efficiency while maintaining comparable performance.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Review
Oncology
Alessandro Allegra et al.
Summary: Multiple myeloma is a complex malignant neoplasm of plasma cells. The application of artificial intelligence, particularly machine learning and deep learning algorithms, has significant implications in the diagnosis and treatment of multiple myeloma.
Article
Engineering, Multidisciplinary
Hong Yang et al.
Summary: In this paper, a novel method of ultra-lightweight convolution neural network (CNN) design based on neural architecture search (NAS) and knowledge distillation (KD) is proposed. It can automatically construct the space target ISAR image recognition model with ultra-lightweight and high accuracy. The effectiveness of the proposed method is verified by simulation experiments on the ISAR image dataset of five types of space targets.
DEFENCE TECHNOLOGY
(2022)
Review
Computer Science, Artificial Intelligence
Meenal V. Narkhede et al.
Summary: Neural network weight initialization is a critical step that directly affects the convergence of the network. This paper discusses various weight initialization techniques, classifies them, and provides an overview of different techniques in the field.
ARTIFICIAL INTELLIGENCE REVIEW
(2022)
Article
Computer Science, Artificial Intelligence
Yair A. Andrade-Ambriz et al.
Summary: A temporal convolutional neural network method is proposed for analyzing and recognizing human activities using spatio-temporal features, achieving improved accuracy in classification results through optimal use of computational resources, and providing real-time classification results.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jinhong Huang et al.
Summary: This study evaluates the influence of image contrast, human anatomy, sampling pattern, undersampling factor, and noise level on the generalization of a deep cascade of convolutional neural network (DC-CNN) for image reconstruction from undersampled k-space data. The results show that reconstruction quality is highly sensitive to sampling pattern, undersampling factor, and noise level, and less sensitive to image contrast. Deviation in human anatomy between training and test data leads to reduction in image quality, particularly for brain datasets. Transfer learning shows potential for image reconstruction from different datasets than those used for training.
MAGNETIC RESONANCE IMAGING
(2022)
Article
Computer Science, Artificial Intelligence
Yirong Xie et al.
Summary: Evolutionary Neural Architecture Search (ENAS) is an automated method for designing deep network architecture, which has gained extensive attention in the field of automated machine learning. This paper proposes improvements in two aspects: the introduction of efficient CNN-based building blocks and a triplet attention mechanism to enhance the effectiveness and classification performance of the generated architectures, and the use of a random forest-based performance predictor to reduce computation in training. Experimental results demonstrate that the proposed algorithm significantly reduces computational resources while achieving competitive classification performance on the CIFAR dataset.
Article
Engineering, Environmental
Andre Zamith Selvaggio et al.
Summary: This paper trained and tested long short-term memory recurrent neural networks on CH4 leakage source in a chemical process module, with models using different values of timesteps to predict the source of leakage. The datasets were obtained through 3D-CFD simulations, considering various leak locations, wind speeds, and wind directions. Results showed improved performance with greater values of timesteps, achieving test accuracy over 95.3% and demonstrating good generalization.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2022)
Article
Computer Science, Information Systems
Jiang Yao et al.
Summary: An automatic labanotation generation algorithm based on deep learning is proposed, which can accurately and quickly record and preserve traditional dance movements. The algorithm performs well in terms of classification and recognition and can create more accurate dance music for simple human movements.
WIRELESS COMMUNICATIONS & MOBILE COMPUTING
(2022)
Article
Computer Science, Information Systems
Fugang Liu et al.
Summary: Based on a comparative analysis of LSTM and GRU networks, this paper optimizes the structure of GRU network and proposes a new modulation recognition method based on feature extraction and deep learning algorithm. The proposed method achieves high recognition rate at low SNR.
TSINGHUA SCIENCE AND TECHNOLOGY
(2022)
Article
Computer Science, Information Systems
Gunju Park et al.
Summary: This paper proposes a NAS technique, called CondNAS, for conditional CNN architecture. By using machine learning models and genetic algorithm, CondNAS efficiently finds a near-optimal conditional CNN architecture. The experimental results show that the conditional CNNs from CondNAS are 2.52 times faster than the CNNs from OFA on GPU and 1.75 times faster on CPU.
Article
Computer Science, Hardware & Architecture
Francesco Daghero et al.
Summary: This study introduces a method for human activity recognition (HAR) on embedded devices using one-dimensional convolutional neural networks (CNNs). By optimizing hyperparameters and applying quantization techniques, efficient CNN models can be obtained, achieving good trade-offs between memory, latency, and energy consumption. The use of adaptive inference allows for a more flexible HAR system, and the proposed CNNs outperform previous deep learning methods in multiple benchmarks.
ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Adi Alhudhaif et al.
Summary: The study developed a reliable convolutional-neural-network (CNN) model for the classification of COVID-19 from chest X-ray views, aiming to address bias issues from the database. The DenseNet-201 architecture outperformed other models with the highest accuracy, precision, recall, and F1-scores.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Chemistry, Analytical
Ge Gao et al.
Summary: This study proposed a human activity classification and recognition model based on smartphone inertial sensor data, which involves feature extraction and recognition on different classifiers. Experimental results showed that dynamic and transitional actions performed well on support vector machines, while static actions had better classification effects on ensemble classifiers.
Article
Chemistry, Analytical
Xiaojuan Wang et al.
Summary: Researchers proposed a method using NAS for HAR task model search, achieving results superior to manually adjusted best models. By using a multi-objective search algorithm and considering the balance between computation speed and complexity, excellent performance was achieved.
Article
Computer Science, Artificial Intelligence
Haoyu Zhang et al.
Summary: The article proposes a computationally efficient framework for evolutionary search of convolutional networks based on a directed acyclic graph, reducing the computational costs of training deep neural networks by using random sampling of parent nodes and a node inheritance strategy, as well as introducing a channel attention mechanism in the search space to enhance feature processing capability. Experimental results show that the algorithm is competitive in terms of computational efficiency and learning performance.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Information Systems
R. Raja Subramaniam et al.
Summary: With the development of modern e-healthcare, the need for ambulatory healthcare for those who are physically or mentally unwell, elderly, or children has become prominent. A hybrid algorithm utilizing deep learning and genetic algorithms, running on fog-assisted cloud framework, demonstrates higher accuracy in inferring human activities compared to state-of-the-art algorithms.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2021)
Article
Geochemistry & Geophysics
Mercedes E. Paoletti et al.
Summary: A new few-parameter CNN for HSI classification, based on shift operations, is introduced to reduce the number of parameters and computational complexity. The method shows promising results in terms of computational performance and classification accuracy.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2021)
Article
Computer Science, Artificial Intelligence
Soumali Roychowdhury et al.
Summary: Deep learning architectures can develop feature representations and classification models in an integrated way during training, and integrating prior knowledge into learning can reduce the amount of required training data.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Linlin Guo et al.
Summary: This study focuses on reducing accuracy differences among individuals in activity recognition in indoor environments, and proposes a novel deep learning model LCED to achieve this goal.
Article
Engineering, Biomedical
Mohamed H. Abdelhafiz et al.
Summary: This paper describes the development of a human gait activity recognition system, which involves simplifying a multi-sensor system to a single sensor system and improving prediction accuracy through sensor selection and algorithm modification.
PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART H-JOURNAL OF ENGINEERING IN MEDICINE
(2021)
Article
Computer Science, Information Systems
Mutegeki Ronald et al.
Summary: Advances in deep learning model design have expanded the scope of applications, including the popular field of human activity recognition (HAR). The proposed iSPLInception model demonstrates superior performance on several public HAR datasets compared to existing deep learning architectures aimed at solving the HAR problem in recent years.
Article
Computer Science, Hardware & Architecture
Mohammad Loni et al.
MICROPROCESSORS AND MICROSYSTEMS
(2020)
Article
Computer Science, Information Systems
Majid Ali Khan Quaid et al.
MULTIMEDIA TOOLS AND APPLICATIONS
(2020)
Article
Chemistry, Multidisciplinary
Mei Chee Leong et al.
APPLIED SCIENCES-BASEL
(2020)
Article
Computer Science, Artificial Intelligence
Yanan Sun et al.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2020)
Article
Engineering, Multidisciplinary
Muhammet Fatih Aslan et al.
Article
Automation & Control Systems
Yanan Sun et al.
IEEE TRANSACTIONS ON CYBERNETICS
(2020)
Article
Computer Science, Software Engineering
D. van Kuppevelt et al.
Article
Computer Science, Information Systems
Abdullah Asim Yilmaz et al.
Article
Computer Science, Information Systems
Hyeungill Lee et al.
Proceedings Paper
Computer Science, Information Systems
Jahandad et al.
FIFTH INFORMATION SYSTEMS INTERNATIONAL CONFERENCE
(2019)
Article
Computer Science, Information Systems
Walaa N. Ismail et al.
Review
Computer Science, Artificial Intelligence
Kenneth O. Stanley et al.
NATURE MACHINE INTELLIGENCE
(2019)
Article
Chemistry, Analytical
Alejandro Baldominos et al.
Article
Computer Science, Information Systems
Lei Wang et al.
Article
Computer Science, Information Systems
M. Baechlin et al.
METHODS OF INFORMATION IN MEDICINE
(2010)
Article
Engineering, Electrical & Electronic
Qijun Zhao et al.
EURASIP JOURNAL ON ADVANCES IN SIGNAL PROCESSING
(2009)
Article
Thermodynamics
Kwang-Tzu Yang
JOURNAL OF HEAT TRANSFER-TRANSACTIONS OF THE ASME
(2008)
Article
Computer Science, Artificial Intelligence
KO Stanley et al.
EVOLUTIONARY COMPUTATION
(2002)
