A5018152658- Advanced Image Processing Techniques
- Advanced Data Compression Techniques
- Image and Signal Denoising Methods
- Advanced Vision and Imaging
- Advanced Statistical Methods and Models
- Psychometric Methodologies and Testing
- Statistical Methods in Clinical Trials
- Statistical Methods and Bayesian Inference
- Cognitive and developmental aspects of mathematical skills
- Education, Achievement, and Giftedness
- Video Coding and Compression Technologies
- Advanced Statistical Modeling Techniques
- Domain Adaptation and Few-Shot Learning
- Cognitive Abilities and Testing
- Multimodal Machine Learning Applications
- Image Processing Techniques and Applications
- Mathematics Education and Teaching Techniques
- Corneal surgery and disorders
- Ocular Surface and Contact Lens
- Corneal Surgery and Treatments
- Pesticide Residue Analysis and Safety
- Advanced Causal Inference Techniques
- Fluid Dynamics and Turbulent Flows
- Educational Strategies and Epistemologies
- Mental Health via Writing
University of South Florida
2015-2025
National Taiwan University
1998-2025
National Defense Medical Center
2024-2025
Tri-Service General Hospital
2024-2025
National Yang Ming Chiao Tung University
2020-2024
Florida College
2024
National Yunlin University of Science and Technology
2021
Providence University
2019
National Cheng Kung University
2002-2018
National Tsing Hua University
2017
Despite the recent success of deep-learning based semantic segmentation, deploying a pre-trained road scene segmenter to city whose images are not presented in training set would achieve satisfactory performance due dataset biases. Instead collecting large number annotated each interest train or refine segmenter, we propose an unsupervised learning approach adapt segmenters across different cities. By utilizing Google Street View and its time-machine feature, can collect unannotated for at...
Various tests to check the homogeneity of variance assumption have been proposed in literature, yet there is no consensus as their robustness when normality does not hold. This simulation study evaluated performance 14 for one-way ANOVA models terms Type I error control and statistical power. Seven factors were manipulated: number groups, average observations per group, pattern sample sizes population variances, maximum ratio, distribution shape, nominal alpha level test variances. Overall,...
In this paper, a CMOS multichannel transceiver (TRX) is proposed for angular identification in automotive car radar applications. Two transmitters (TXs) and six receivers (RXs) are placed on the same die using 65-nm technology with chip size 3.5 × 3 mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . To generate millimeter-wave (MMW) frequency-modulated continuous-wave signal, an injection-lock frequency sextupler cascaded 1-to-8...
Background Multiple-choice item (MCI) assessments are burdensome for instructors to develop. Artificial intelligence (AI, e.g., ChatGPT) can streamline the process without sacrificing quality. The quality of AI-generated MCIs and human experts is comparable. However, whether equally good across various domain- task-specific prompts remains be determined. Therefore, we ask AI generate high-quality assess learning outcomes from a psychology textbook chapter reading. Objective In an exploratory...
A simulation study is designed to explore the accuracy of attribute parameter estimation in crossed random effects linear logistic test model (CRELLTM) with impact Q-matrix misspecification on using SAS® GLIMMIX procedure a scaling constraint item parameter. In addition, interactions other manipulated factors, such as population distribution, sample size, and density, also investigated. The results indicated that type percent have considerable bias root mean squared error estimates,...
Abstract Human hepatocellular carcinoma (HCC) is the most common liver cancer and third leading cause of cancer‐related deaths worldwide. HCC a malignant tumor that can lead to intrahepatic extrahepatic metastases. Intercellular adhesion molecule 1 (ICAM‐1) involved in metastasis. ICAM‐1 enhances cell‐cell interactions by promoting facilitating cell movement within extracellular matrix. Moreover, more abundant cancerous hepatocytes than non‐cancerous ones. Chemokine (C‐X‐C motif) ligand...
In this article, the authors investigated relations of students' perceptions teachers' oral feedback with teacher expectancies and student self-concept. A sample 1,598 Taiwanese children in Grades 3 to 6 completed measures school Homeroom teachers identified students for whom they had high or low expectancies. Discriminant analysis indicated positive negative academic were more important than nonacademic predicting 2-way multivariate variance showed that boys perceived did girls, fifth-grade...
In this paper, we investigate the emotion recognition ability of pre-training language model, namely BERT. By nature framework BERT, a two-sentence structure, adapt BERT to continues dialogue prediction tasks, which rely heavily on sentence-level context-aware understanding. The experiments show that by mapping into causal utterance pair, is constructed and reply utterance, models can better capture emotions utterance. present method has achieved 0.815 0.885 micro F1 score in testing dataset...
This work aims for transferring a Transformer-based image compression codec from human perception to machine without fine-tuning the codec. We propose transferable framework, termed TransTIC. Inspired by visual prompt tuning, TransTIC adopts an instance-specific generator inject prompts encoder and task-specific decoder. Extensive experiments show that our proposed method is capable of base various tasks outperforms competing methods significantly. To best knowledge, this first attempt...
Cognitive diagnosis models (CDMs) estimate student ability profiles using latent attributes. Model fit to the data needs be ascertained in order determine whether inferences from CDMs are valid. This study investigated usefulness of some popular model statistics detect CDM including relative indices (AIC, BIC, and CAIC), absolute (RMSEA2, ABS(fcor) MAX(χ2jj′)). These were assessed under different settings with respect Q-matrix misspecification misspecification. Results showed that selected...
This work addresses continuous space-time video super-resolution (C-STVSR) that aims to up-scale an input both spatially and temporally by any scaling factors. One key challenge of C-STVSR is propagate information among the frames. To this end, we introduce a local implicit neural function. It has striking feature learning forward motion for continuum pixels. We motivate use from perspective individual trajectories, as opposed mixture trajectories with backward motion. ease interpolation,...
Abstract Corneal wound healing is one of the major issues in ocular surface reconstruction and diseases. Amniotic membrane (AM) transplantation an excellent treatment modality to promote corneal treat It interesting valuable search for another synthetic biocompatible substitute study mechanism AM disorders. Chitosan, second‐most abundant polymer nature, has many biological advantages such as biocompatibility, biodegradability, hemostatic activity, wound‐healing property be used biomedical...
This paper demonstrates a successive second-order intermodulation feed-forward cancelling path, which can reduce the third-order distortion (IMD3) of 24 GHz cascode power amplifier (PA) effectively. By adopting this technique, P <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sat</inf> , OP xmlns:xlink="http://www.w3.org/1999/xlink">1dB</inf> and associated PAE are 16.6 dBm, 14.6 12.6%, respectively. The sweet-spot locates at 5 dBm output with...
This paper proposes a transformer-based learned image compression system. It is capable of achieving variable-rate with single model while supporting the region-of-interest (ROI) functionality. Inspired by prompt tuning, we introduce generation networks to condition autoencoder compression. Our generate content-adaptive tokens according input image, an ROI mask, and rate parameter. The separation mask parameter allows intuitive way achieve coding simultaneously. Extensive experiments...
As an intense laser pulse propagates through underdense plasma, the strong ponderomotive force pushes away electrons and produces a trailing plasma bubble. In meantime itself undergoes extreme nonlinear evolution that results in spectral broadening toward long-wavelength side. By experiment we demonstrate this process can be utilized to generate ultrashort midinfrared pulses with energy three orders of magnitude larger than produced by crystal-based optics. The infrared is encapsulated...
Abstract In this study, biomaterials with different hydrophobic properties including polyvinyl alcohol (PVA), poly(ethylene‐ co ‐vinyl alcohol) (EVAL), tissue culture polystyrene (TCPS), and polyvinylidene fluoride (PVDF) were examined in the bovine corneal endothelial cells (BCECs) system to elucidate their possible impact on clinical demand scientific interest. It was found that BCECs inhibited attach onto PVA surface. Conversely, relatively more EVAL, TCPS, PVDF successfully initiate BCEC...
Previous studies have shown that method effects associated with item wording produce artifactual factors and threaten scale validity. This study examines on a of attitudes toward learning mathematics for Taiwanese U.S. samples. Analyses from series CFA (confirmatory factor analysis) models support the presence both In addition, findings show students tended to report higher means not only substantive but also factor, compared students. The effect sizes mean differences are medium large.
This study was an empirically substantive examination of mathematics achievement differences between urban and rural Taiwanese students at a cognitive attribute level. Participants were eighth-grade who participated in the Trends International Mathematics Science Study 1999. The rule-space method applied to produce diagnostic description students' knowledge, abilities, skills related TIMSS items. results indicated that schools performed better than those on high-level contents (Algebra...
This paper addresses the video rescaling task, which arises from needs of adapting spatial resolution to suit individual viewing devices. We aim jointly optimize downscaling and upscaling as a combined task. Most recent studies focus on image-based solutions, do not consider temporal information. present two joint optimization approaches based invertible neural networks with coupling layers. Our Long Short-Term Memory Video Rescaling Network (LSTM-VRN) leverages information in low-resolution...
As with any test administered across linguistically and culturally diverse groups, evidence suggesting the equivalence of score meaning countries is needed for valid comparisons. The current study examines cross-cultural interpretations from Trends in International Mathematics Science Study (TIMSS)-1999 a cognitive-psychometric perspective. A cognitive model describing knowledge, strategies, processing skills measured by TIMSS-R mathematics was previously validated several countries. In...