A5016753476- Neural Networks and Applications
- Quantum Computing Algorithms and Architecture
- Electronic and Structural Properties of Oxides
- Dental Radiography and Imaging
- Surgical Simulation and Training
- Supercapacitor Materials and Fabrication
- Advancements in Semiconductor Devices and Circuit Design
- Dental Anxiety and Anesthesia Techniques
- Transcranial Magnetic Stimulation Studies
- Nasal Surgery and Airway Studies
- Fractal and DNA sequence analysis
- Natural Language Processing Techniques
- Advancements in Solid Oxide Fuel Cells
- Nutritional Studies and Diet
- Cognitive Science and Education Research
- Text and Document Classification Technologies
- Airway Management and Intubation Techniques
- Advanced Condensed Matter Physics
- Physics of Superconductivity and Magnetism
- Foreign Body Medical Cases
- Magnetic and transport properties of perovskites and related materials
- Thin-Film Transistor Technologies
- Semiconductor materials and devices
- Inorganic Chemistry and Materials
- Neural and Behavioral Psychology Studies
Chang Gung University
2025
National Taiwan University
2022-2023
Max Planck Institute for Chemical Physics of Solids
2021-2022
National Yang Ming Chiao Tung University
2005-2022
Academia Sinica
2004
University of Science and Technology of China
2000
Food classification is the foundation for developing food vision tasks and plays a key role in burgeoning field of computational nutrition. Due to complexity requiring fine-grained classification, recent academic research mainly modifies Convolutional Neural Networks (CNNs) and/or Vision Transformers (ViTs) perform category classification. However, learn features, CNN backbone needs additional structural design, whereas ViT, containing self-attention module, has increased complexity. In...
Abstract Background Rhinolaryngoscopes are one of the most widely used tools by otolaryngologists and speech-language pathologists in current clinical practice. However, there is limited data on adverse events associated with or caused use rhinolaryngoscopes. Objective In this study, we Manufacturer User Facility Device Experience (MAUDE) database aim providing insights that may assist better understanding limitations these devices selecting appropriate procedures for their specific setting....
We propose an approach that uses mirror point pairs and a multiple classifier system to reduce the classification time of support vector machine (SVM). Decisions made with simple classifiers formed from are integrated approximate rule single SVM. A coarse-to-fine is developed for selecting given number member classifiers. clustering method, derived similarities between classifiers, used coarse selection. greedy strategy then fine selection Selected further refined by finding weighted...
This brief studies the electrical characteristics of a series polysilicon thin-film transistors (poly-Si TFTs) with different numbers multiple channels various widths, lightly doped drain (LDD) structures. The nanoscale TFT ten 67-nm-wide split (M10) has superior and more uniform than other TFTs. Additionally, experimental results reveal that performance proposed TFTs enhances each channel width decreasing, yielding profile from single-gate to tri-gate structure.
Electroencephalography (EEG) is a critical tool in neuroscience and clinical practice for monitoring analyzing brain activity. Traditional neural network models, such as EEGNet, have achieved considerable success decoding EEG signals but often struggle with the complexity high dimensionality of data. Recent advances quantum computing present new opportunities to enhance machine learning models through (QML) techniques. In this paper, we introduce Quantum-EEGNet (QEEGNet), novel hybrid that...
Dense oxygen permeable membranes are expected to have a great impact on production and oxygen-involved industrial processes. This work demonstrates that Bi2Sr2-CaCu2O8, well-known superconductor, is very at intermediate temperatures while remaining impervious nitrogen other gases. The Figure shows possible mechanism for the process.
ReO${}_{3}$ is a remarkable transition metal oxide in that it has the highest conductivity of all oxides, comparable even with silver. Using state art bulk-sensitive angle-resolved photoelectron spectroscopy, authors are able to observe clear dispersions Re 5$d$ and O 2$p$ derived bands as well momentum splitting Fermi surface due spin-orbit interaction. The experimental results compared quantitatively density functional theory band structure calculations, thereby providing deeper...
Orientation dependent hard x-ray photoelectron spectroscopy can make direct images of the orbitals making up band structure. The imaging provides energy specific information in real-space and is calculation-free.
Decoding images from non-invasive electroencephalographic (EEG) signals has been a grand challenge in understanding how the human brain process visual information real-world scenarios. To cope with issues of signal-to-noise ratio and nonstationarity, this paper introduces MUltimodal Similarity-keeping contrastivE learning (MUSE) framework for zero-shot EEG-based image classification. We develop series multivariate time-series encoders tailored EEG assess efficacy regularized contrastive...
In this paper, we propose a novel framework for multimodal contrastive learning utilizing quantum encoder to integrate EEG (electroencephalogram) and image data. This groundbreaking attempt explores the integration of encoders within traditional framework. By leveraging unique properties computing, our method enhances representation capabilities, providing robust analyzing time series visual information concurrently. We demonstrate that effectively captures intricate patterns signals...
NECOMIMI (NEural-COgnitive MultImodal EEG-Informed Image Generation with Diffusion Models) introduces a novel framework for generating images directly from EEG signals using advanced diffusion models. Unlike previous works that focused solely on EEG-image classification through contrastive learning, extends this task to image generation. The proposed NERV encoder demonstrates state-of-the-art (SoTA) performance across multiple zero-shot tasks, including 2-way, 4-way, and 200-way, achieves...
The Psycho Frame, a sophisticated system primarily used in Universal Century (U.C.) series mobile suits for NEWTYPE pilots, has evolved as an integral component harnessing the latent potential of mental energy. Its ability to amplify and resonate with pilot's psyche enables real-time control, creating unique applications such psychomagnetic fields sensory-based weaponry. This paper presents development novel robotic control inspired by combining electroencephalography (EEG) deep learning...
Background: During laparoscopic surgery, surgical gauze is usually inserted into body cavity to help achieve hemostasis. Retention of in the may necessitate reoperation and increase risk. Using deep learning technology, this study aimed propose a neural network model for detection from video record prescence gauze.Methods: A total 3000 images obtained records colectomy/proctectomy were prepared as dataset. The dataset was divided training group (1250 images) testing (1750 images). trained by...