A5101566996- Fault Detection and Control Systems
- Neural Networks and Applications
- Video Surveillance and Tracking Methods
- EEG and Brain-Computer Interfaces
- Fire Detection and Safety Systems
- Anomaly Detection Techniques and Applications
- Currency Recognition and Detection
- Diverse Aspects of Tourism Research
- Advanced Sensor and Control Systems
- Smart Grid Security and Resilience
- Forecasting Techniques and Applications
- Electrical Fault Detection and Protection
- Adaptive Control of Nonlinear Systems
- Genomics and Phylogenetic Studies
- Neural Networks and Reservoir Computing
- Handwritten Text Recognition Techniques
- Generative Adversarial Networks and Image Synthesis
- Brain Tumor Detection and Classification
- AI in cancer detection
- Advanced Algorithms and Applications
- Elevator Systems and Control
- Soft Robotics and Applications
- Blind Source Separation Techniques
- Radiomics and Machine Learning in Medical Imaging
- Sleep and Work-Related Fatigue
Shanghai Jiao Tong University
2020-2024
Engineering Systems (United States)
2023
University of Memphis
2023
Institute of Electrical and Electronics Engineers
2023
Antea Group (France)
2023
Ministry of Education of the People's Republic of China
2022
This article presents a new aviation brain-computer interface, which includes the construction of color brain power map and cognitive detection network. The developed network, Bpmnet, can effectively detect state brain. To improve effectiveness model parameter optimization algorithms, momentum batch normalization are proposed during Bayesian posterior inference. Bpmnet reduces risk overfitting increases uncertainty outlier prediction. Experimental results demonstrate that our approach...
This work proposes a nonparametric prior induced deep sum-logarithmic-multinomial mixture (DSLMM) model to detect pilots' cognitive states through the developed brain power map. DSLMM uses multinormal distribution infer latent variable of each neuron in first layer network. These variables obeyed sum-logarithmic that is backpropagated its observation vector and number neurons next layer. Multinormal used segment extended form matrix associated with width also an adaptive topic-layer...
While various biomimetic robotic haptic sensors are often utilized to measure multifarious physical interactions identify material properties under human exploratory procedures (EPs), traditional methods unable fuse well multimodal measurements EPs. In order solve this problem, an innovative hybrid joint group kernel sparse coding model for recognition EPs is proposed. First, a series of feature representations according the characteristics different measures introduced. Second, we propose...
To improve the convergence rate of path-following errors a bionic snake robot and reduce overshoot kinematic errors, control method based on fuzzy line-of-sight (LOS) guidance for with unknowns is developed. This designs time-varying LOS forward distance sliding mode switching gain according to rules, offsets external interferences, weakens damage chattering joints, reduces negative impact interferences path-following. In addition, uncertain friction coefficients are predicted by using...
As a group of complex neurodevelopmental disorders, autism spectrum disorder (ASD) has been reported to have high overall prevalence, showing an unprecedented spurt since 2000. Due the unclear pathomechanism ASD, it is challenging diagnose individuals with ASD merely based on clinical observations. Without additional support biochemical markers, difficulty diagnosis could impact therapeutic decisions and, therefore, lead delayed treatments. Recently, accumulating evidence shown that both...
Crowd counting is considered as the essential computer vision application that uses convolutional neural network to model crowd density regression task. However, vision-based models are hard extract feature under low-quality conditions. As we know, visual and audio used widely media platforms for human beings touch physical change of world. The cross-modal information gives us an alternative method solving In this case, in order solve problem, a named Audio-Visual Multi-Scale Network (AVMSN)...
In the context of Industry 4.0, medical industry is horizontally integrating resources entire through Internet Things (IoT) and digital interconnection technologies. Speeding up establishment public retrieval database diagnosis-related historical data a common call for industry. Among them, Magnetic Resonance Imaging (MRI) system, which one key tools secure private Medical (IoMT), significant patients to check their conditions doctors make clinical diagnoses securely privately. Hence, this...
Recognition of amount in figures the financial multi-bill scenes is crucial for automatic banking business. However, diversity business and limitation customer data privacy determine that it difficult to collect a large number sample datasets. Aiming at problem insufficient training low accuracy detection model, this article proposes new generative adversarial network (GAN) generate samples expand bill dataset, which then adopted train framework recognition amount. In proposed WGAN-SA,...
The Backpropagation Multidimensional Taylor Network (BP-MTN) classifier improved based on MTN may face issues such as overfitting and gradient vanishing when solving complex classification problems. To address these issues, this paper proposes a novel classifier, the Residual (ResMTN) classifier. introduces idea of (ResNet) in MTN, by adding direct connection between input fully connected layer beside polynomial that is, using skip to layer, thus improving generalization ability ResMTN. In...