A5100411058- Climate variability and models
- Natural Language Processing Techniques
- Hydrology and Watershed Management Studies
- Advanced Optical Sensing Technologies
- Topic Modeling
- Cryospheric studies and observations
- Speech Recognition and Synthesis
- Plant Water Relations and Carbon Dynamics
- Infrared Target Detection Methodologies
- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- Remote Sensing and LiDAR Applications
- Advanced Measurement and Detection Methods
- Advanced Image Fusion Techniques
- Remote Sensing and Land Use
- Image Processing Techniques and Applications
- Optical measurement and interference techniques
- Climate change and permafrost
- Meteorological Phenomena and Simulations
- Flood Risk Assessment and Management
- Photoacoustic and Ultrasonic Imaging
- Hydrology and Drought Analysis
- Optical Systems and Laser Technology
- Arctic and Antarctic ice dynamics
- Anomaly Detection Techniques and Applications
National Taiwan Normal University
2023-2025
University of Alberta
2025
Xi'an University of Science and Technology
2024
National University of Defense Technology
2018-2024
China Institute of Water Resources and Hydropower Research
2009-2024
Zhejiang Chinese Medical University
2024
Beijing Institute of Fashion Technology
2024
Texas A&M University
2024
State Key Laboratory of Pulsed Power Laser Technology
2020-2024
Northeast Agricultural University
2024
Reading comprehension QA tasks have seen a recent surge in popularity, yet most works focused on fact-finding extractive QA. We instead focus more challenging multi-hop generative task (NarrativeQA), which requires the model to reason, gather, and synthesize disjoint pieces of information within context generate an answer. This type multi-step reasoning also often understanding implicit relations, humans resolve via external, background commonsense knowledge. first present strong baseline...
The European Center for Medium-Range Weather Forecasts (ECMWF) released its latest reanalysis dataset named ERA5 in 2017. To assess the performance of Antarctica, we compare near-surface temperature data from and ERA-Interim with measured 41 weather stations. has a strong linear relationship monthly observations, statistical significant correlation coefficients (p < 0.05) are higher than 0.95 at all stations selected. shows regional differences, correlations high West Antarctica low East...
In this paper, a novel deep architecture named BraidNet is proposed for person re-identification. has specially designed WConv layer, and the cascaded structure learns to extract comparison features of two images, which are robust misalignments color differences across cameras. Furthermore, Channel Scaling layer optimize scaling factor each input channel, helps mitigate zero gradient problem in training phase. To solve imbalanced volume negative positive samples, Sample Rate Learning...
Success in natural language inference (NLI) should require a model to understand both lexical and compositional semantics. However, through adversarial evaluation, we find that several state-of-the-art models with diverse architectures are over-relying on the former fail use latter. Further, this compositionality unawareness is not reflected via standard evaluation current datasets. We show removing RNNs existing or shuffling input words during training does induce large performance loss...
Abstract The estimation of reference evapotranspiration (ET0) is important in hydrology research, irrigation scheduling design and water resources management. This study explored the capability eight machine learning models, i.e., Artificial Neuron Network (ANN), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), Extreme (XGBoost), Multivariate Adaptive Regression Spline (MARS), Support Vector Machine (SVM), Learning a novel Kernel-based Nonlinear Extension Arps Decline (KNEA)...
Increasing economy and population result in the rapid growth of solid waste China, with more than 70% being carbonaceous components. The generated carbon-bearing is called as organic (OSW). In this Review, OSWs from both municipal industrial sources are discussed divided into (MSW), sludge, polymer waste, biomass-like according to their variations origins characteristics China. Agricultural (i.e., biomass) large-scale (e.g., coal waste) not considered, because well-established fundamentals....
Food safety is closely related to human health. However, the regulation and testing processes for food are intricate resource-intensive. Therefore, it necessary address risks using a combination of deep learning, Internet Things, smartphones, quick response codes, smart packaging, other technologies. Intelligent designs that combine digital systems advanced functionalities with biosensors hold great promise revolutionizing current practices. This review introduces concept Safety 4.0,...
This paper presents a privacy-preserving federated learning (PPFL) system for the detection of android malware. The proposed PPFL allows mobile devices to collaborate together training classifier without exposing sensitive information, such as application programming interface (API) calls and permission configuration, learned local model by each device. work implements based on support vector machine (SVM) secure multi-party computation techniques. It also demonstrates feasibility using...
This article proposes a zero-shot learning-based framework for light field depth estimation, which learns an end-to-end mapping solely from input to the corresponding disparity map with neither extra training data nor supervision of groundtruth depth. The proposed method overcomes two major difficulties posed in existing methods and is thus much more feasible practice. First, it saves huge burden obtaining variety scenes serve as labels during training. Second, avoids severe domain shift...
Depth estimation is an important computer vision problem with many practical applications to mobile devices. While solutions have been proposed for this task, they are usually very computationally expensive and thus not applicable on-device inference. To address problem, we introduce the first Mobile AI challenge, where target develop end-to-end deep learning-based depth that can demonstrate a nearly real-time performance on smartphones IoT platforms. For this, participants were provided new...
Abstract The polar amplification (PA) has become the focus of climate change. However, there are seldom comparisons among Earth’s three poles Arctic (latitude higher than 60 °N), Antarctica (Antarctic Ice Sheet) and Third Pole (the High Mountain Asia with elevation 4000 m) under different socioeconomic scenarios. Based on CMIP6 multi-model ensemble, two types PA index (PAI) have been defined to quantify intensity variations, PAI1/PAI2 is as ratio absolute value surface air temperature linear...
Tantalum nitride is widely considered as a promising photoanode material for its suitable band structure well the high theoretical conversion efficiency in solar water splitting. However, it limited to inefficient photoinduced electron–hole pair separation and interfacial dynamics photoelectrochemical oxygen evolution reaction. Herein, multiple layers including Ti x Si y NiFeCoO were fabricated based on engineering regulate tandem electric states efficient transfer of energy carriers....
This study explores the relationship between physical pain sensitivity and word meaning perception, building on findings by Reuter et al. (2016) that individuals experiencing higher levels of associate words more strongly with pain. Grounded in Body-Specificity Hypothesis (BSH; Casasanto, 2009), research aims to elucidate mechanisms which influences cognitive interpretations language. A pharmaceutical intervention utilizing an analgesic will be employed investigate these effects.
In 2017, we called it "WeChat Mini Program and the first year". Since then, development of WeChat has been booming. At present, next generation mobile Internet industry is developing toward "micro, light, small", like Program, this revolutionary lightweight "APP" that does not require downloading, runs out, simultaneously information publishing, advertising, service functions, will become best choice to replace client APP. This article analyze system implementation prospects study...