- Video Analysis and Summarization
- Photonic Crystals and Applications
- Recommender Systems and Techniques
- Advanced Memory and Neural Computing
- Neural dynamics and brain function
- Online Learning and Analytics
- Optical Coatings and Gratings
- Multimodal Machine Learning Applications
- Sports Analytics and Performance
- Sports Dynamics and Biomechanics
- Radio Frequency Integrated Circuit Design
- Metamaterials and Metasurfaces Applications
- Anomaly Detection Techniques and Applications
- Artificial Intelligence in Healthcare
- Advanced Photonic Communication Systems
- Cell Image Analysis Techniques
- CCD and CMOS Imaging Sensors
- Radiomics and Machine Learning in Medical Imaging
- Information Technology Governance and Strategy
- Machine Learning in Healthcare
- Analog and Mixed-Signal Circuit Design
- BIM and Construction Integration
- Optical Network Technologies
- Big Data and Business Intelligence
- Medical Imaging Techniques and Applications
Kaiser Permanente
2024
Hanoi University of Science and Technology
2017-2022
East Asia School of Theology
2022
North Carolina State University
2015
University of California, Los Angeles
2012
University of Kaiserslautern
2004-2007
Job openings often go unfulfilled despite a surfeit of unemployed or underemployed workers. One the main reasons for this is mismatch between skills required by employers and that workers possess. This mismatch, also known as gap, can pose socio-economic challenges an economy. A first step in alleviating gap to accurately detect human capital data such resumes job ads. Comprehensive accurate detection facilitates analysis labor market dynamics. It helps bridge divide supply demand...
The online recruitment matching system has been the core technology and service platform in CareerBuilder. One of major challenges an scenario is to provide good matches between job posts candidates using a recommender on scale. In this paper, we discussed techniques for applying embedding-based large scale matching. To learn comprehensive effective embedding candidates, have constructed fused-embedding via different levels representation learning from raw text, semantic entities location...
Major League Baseball, a professional baseball league in the US and Canada, is one of most popular sports leagues world. Partially because its popularity wide availability data from games, has become subject significant statistical mathematical analysis. Pitch analysis especially useful for helping team better understand pitch behavior it may face during game, allowing to develop corresponding batting strategy combat predicted behavior. We apply several common machine learning classification...
Image classification is always an interesting problem due to its practical applications in real life. With a capability of self-learning features, modern Convolution Neural Network (CNN) models can achieve high accuracy on large and complex benchmark datasets. However, their computation costs, the CNN experience energy consumption problems during training implementation hardware which limits utilisation mobile embedded applications. Recently, Spiking (SNN) has been proposed overcome...
The phenomena of beam-steering and negative ray refraction in a graded 2D photonic crystal structure interface with plane tapered transitions to homogeneous waveguide regions is used for efficient far-field refocusing highly divergent Gaussian beams. It shown by numerical simulation that 14microm long able refocus small-spot beam an efficiency about 90%.
Confocal fluorescence microscopy is one of the most accessible and widely used imaging techniques for study biological processes. Scanning confocal allows capture high-quality images from 3D samples, yet suffers well-known limitations such as photobleaching phototoxicity specimens caused by intense light exposure, which limits its use in some applications, especially living cells. Cellular damage can be alleviated changing parameters to reduce often at expense image quality. Machine/deep...
Negative ray refraction in optimized nonuniform triangular 2D photonic crystal slabs is used for efficient far-field refocusing of a point source. It shown by numerical simulations that with thicknesses between 10 and 0.6μm conjunction proposed superstructure transitions are able to refocus source diffraction limited efficiencies up 97%.
Control and manipulation of properties - such as vacuum modal confinement, spatial harmonic content, phase velocity, group velocity are reasons why an all-dielectric beam-driven accelerator would be expected to benefit by borrowing from the field optical bandgap photonics. We outline general conditions for coherent Cerenkov radiation in a photonic crystal, illustrated three examples: two Bragg mirrors separated gap, woodpile with linear waveguide channel, stack slabs having planar channel.
One-, two- or three-dimensional periodic structures with an omni-directional photonic bandgap are well known as "photonic crystals". For applications in new improved functional components used for communications we currently investigated more detail their spatial and chromatic dispersion. Especially the strong angular difference between phase group velocity vicinity of optical may be practical devices. These strongly frequency dependent beam-steering effects were demonstrated by one author...
This paper presents at high speed and wideband direct receiver based on delta-sigma architecture. The effects of inter-symbol inference are discussed in this paper. In addition, a robust technique using Decision Feedback Equalizer for multi-level (three bits) non-return-to-zero feedback Digital-to-Analog Converter the is also studied. proposed to reduce dynamic error overall system Signal-to-Noise-and Distortion-Ratio, as well compensate non-idealities Digital-to- Analog multi-Gb/s data rate...
Job transitions and upskilling are common actions taken by many industry working professionals throughout their career. With the current rapidly changing job landscape where requirements constantly sectors emerging, it is especially difficult to plan navigate a predetermined career path. In this work, we implemented system automate collection classification of training videos help seekers identify acquire skills necessary transition next step in We extracted educational built machine...