A5034115794- Silicon Carbide Semiconductor Technologies
- Semiconductor materials and devices
- Advancements in Semiconductor Devices and Circuit Design
- Radiomics and Machine Learning in Medical Imaging
- Advanced Neural Network Applications
- AI in cancer detection
- COVID-19 diagnosis using AI
- Advanced Vision and Imaging
- Brain Tumor Detection and Classification
- Electrostatic Discharge in Electronics
- Optical measurement and interference techniques
- Remote Sensing and LiDAR Applications
- Medical Image Segmentation Techniques
- Radio Frequency Integrated Circuit Design
- Wildlife Ecology and Conservation
- Animal Vocal Communication and Behavior
- Metal Extraction and Bioleaching
- Minerals Flotation and Separation Techniques
- Thin-Film Transistor Technologies
- Lung Cancer Diagnosis and Treatment
- Semiconductor materials and interfaces
- Integrated Circuits and Semiconductor Failure Analysis
- Species Distribution and Climate Change
- Handwritten Text Recognition Techniques
- Advanced DC-DC Converters
University of Lincoln
2011-2023
Fuzhou University
2022-2023
Infrastructure University Kuala Lumpur
2022
Shangluo University
2022
Taiwan Semiconductor Manufacturing Company (China)
2020
NEC (Japan)
2013
Xi'an University of Technology
2013
Hua Hong Semiconductor (China)
2012
University of Sheffield
2009-2011
Deep learning architecture with convolutional neural network achieves outstanding success in the field of computer vision. Where U-Net has made a great breakthrough biomedical image segmentation and been widely applied wide range practical scenarios. However, equal design every downsampling layer encoder part simply stacked convolutions do not allow to extract sufficient information features from different depths. The increasing complexity medical images brings new challenges existing...
This paper presents a supervised texton based approach for the accurate segmentation and measurement of ultrasound fetal head (BPD, OFD, HC) femur (FL). The method consists several steps. First, non-linear diffusion technique is utilized to reduce speckle noise. Then, on assumption that cross sectional intensity profiles skull can be approximated by Gaussian-like curves, multi-scale multi-orientation filter bank designed extract features specific anatomic structure. extracted cues, together...
Abstract There exist various types of information on retail food packages, including use by date, product name and so on. The correct coding dates packages is vitally important for avoiding potential health risks to customers caused erroneous mislabelling dates. It extremely tedious laborious check the manually a human operator, which prone generate errors thus an automatic system validating correctness needed. In order construct such system, firstly it needs correctly recognize packages....
Chest radiographs are the most commonly performed radiological examinations for lesion detection. Recent advances in deep learning have led to encouraging results various thoracic disease detection tasks. Particularly, architecture with feature pyramid network performs ability recognise targets different sizes. However, such networks difficult focus on regions chest X-rays due their high resemblance vision. In this paper, we propose a dual attention supervised module multi-label radiographs,...
Shallow Trench Isolation (STI) and Stepped Oxide (SO) -based N-type LDMOS are simultaneously studied for the first time low-voltage power device in this paper. Ultra-low specific on-resistance (Rsp) can be obtained both STI- SO-based by scaling channel length (LCH) as well using optimized drift implant. It was indicated that, to some extent, LCH effectively reduces Rsp without sacrificing breakdown voltage (BV). This is mainly ascribed enhanced conducting capability of shorter cell pitch....
This paper presents a 500 V high voltage NLDMOS with breakdown (<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M1"><mml:mrow><mml:msub><mml:mrow><mml:mi>V</mml:mi></mml:mrow><mml:mrow><mml:mtext>B</mml:mtext><mml:mtext>D</mml:mtext></mml:mrow></mml:msub></mml:mrow></mml:math>) improved by field plate technology. Effect of metal (MFP) and polysilicon (PFP) on improvement is studied. The coeffect MFP PFP drain side has also been investigated. A demonstrated 37 μ m drift length...
A 0.35μm 600V ultra-thin epitaxial BCD technology for high voltage gate driver IC is developed in this work, including LDMOS, isolation ring, asymmetric 20V NMOS, PMOS, symmetric PMOS and BJT. Only 15 photo layers are used the proposed 1P2M technology. The experimental devices demonstrate that BVs of divided RESURF structures 770V, between V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</sub> LDMOS side...
This paper presents the analysis and derivation of geometric relation between vanishing points camera parameters central catadioptric systems. These correspond to three mutually orthogonal directions 3D real world coordinate system (i.e. X, Y Z axes). Compared (VPs) in perspective projection, advantages VPs under projection are that there normally two for each set parallel lines, since lines projected conics image plane. Also, their usually located inside frame. We show knowledge...
Two novel implementations (N+&P+-Polysilicon Memory/Memory1 and Memory-First/Memory2) a conventional way (Memory-Last/Memory3) for low power self-aligned 1.5-T SONOS are demonstrated in this paper. The process flows steps of those devices presented characteristic them compared investigated. It was founded that electrical characteristics Memory1/Memory3 is much superior to Memory2. degeneration Memory2 mainly ascribed the repeated doped channel source/drain regions. Moreover, Memory1 Memory3...
Bird populations are an important bio-indicator; so collecting reliable data is useful for ecologists helping conserve and manage fragile ecosystems.However, existing manual monitoring methods labour-intensive, time-consuming, error-prone.The aim of our work to develop a system, capable automatically classifying individual bird species in flight from videos.This challenging, but appropriate use the field, since there often requirement identify flight, rather than when stationary.We present...
Deep learning architecture with convolutional neural network (CNN) achieves outstanding success in the field of computer vision. Where U-Net, an encoder-decoder structured by CNN, makes a great breakthrough biomedical image segmentation and has been applied wide range practical scenarios. However, equal design every downsampling layer encoder part simply stacked convolutions do not allow U-Net to extract sufficient information features from different depths. The increasing complexity medical...
Domain-generalized nuclei segmentation refers to the generalizability of models unseen domains based on knowledge learned from source and is challenged by various image conditions, cell types, stain strategies. Recently, Segment Anything Model (SAM) has made great success in universal interactive prompt modes (e.g., point box). Despite its strengths, original SAM presents limited adaptation medical images. Moreover, requires providing manual bounding box prompts for each object produce...