A5101762391- High Altitude and Hypoxia
- Neuroscience of respiration and sleep
- Pulmonary Hypertension Research and Treatments
- Social Media and Politics
- COVID-19 Clinical Research Studies
- Topic Modeling
- Chronic Obstructive Pulmonary Disease (COPD) Research
- Advanced Neural Network Applications
- Heart Rate Variability and Autonomic Control
- Domain Adaptation and Few-Shot Learning
- Cardiovascular Function and Risk Factors
- Cardiovascular and Diving-Related Complications
- Advanced Image and Video Retrieval Techniques
- Gout, Hyperuricemia, Uric Acid
- Travel-related health issues
- Hate Speech and Cyberbullying Detection
- Cardiovascular Effects of Exercise
- Long-Term Effects of COVID-19
- Multimodal Machine Learning Applications
- Speech Recognition and Synthesis
- RNA modifications and cancer
- COVID-19 and healthcare impacts
- Opinion Dynamics and Social Influence
- Brain Tumor Detection and Classification
- Sociopolitical Dynamics in Russia
Virginia Tech
2023-2025
Shanghai Jiao Tong University
2025
Hubei University of Chinese Medicine
2025
Tencent (China)
2024
Northeast Normal University
2023
Chinese PLA General Hospital
2015-2022
Army Medical University
2014-2022
Zhejiang University
2022
Army Medical College
2022
82th Hospital of Pla
2020-2022
We revisit large kernel design in modern convolutional neural networks (CNNs). Inspired by recent advances vision transformers (ViTs), this paper, we demonstrate that using a few kernels instead of stack small could be more powerful paradigm. suggested five guidelines, e.g., applying re-parameterized depthwise convolutions, to efficient high-performance large-kernel CNNs. Following the propose RepLKNet, pure CNN architecture whose size is as 31×31, contrast commonly used 3×3. RepLKNet...
Abstract Hypertension is proved to be associated with severity and mortality in coronavirus disease 2019 (COVID‐19). However, little known about the effects of pre‐admission and/or in‐hospital antihypertension treatments on clinical outcomes. Thus, this study aimed investigate association between blood pressure (BP) control COVID‐19–related outcomes compare different treatments. This included 2864 COVID‐19 patients 1628 were hypertensive. Patients grouped according their BP during...
We introduce a multi-step reasoning framework using prompt-based LLMs to examine the relationship between social media language patterns and trends in national health outcomes. Grounded fuzzy-trace theory, which emphasizes importance of "gists" causal coherence effective communication, we Role-Based Incremental Coaching (RBIC), LLM framework, identify gists at-scale. Using RBIC, systematically extract from subreddit discussions opposing COVID-19 measures (Study 1). then track how these...
The well-designed structures in neural networks reflect the prior knowledge incorporated into models. However, though different models have various priors, we are used to training them with model-agnostic optimizers such as SGD. In this paper, propose incorporate model-specific by modifying gradients according a set of hyper-parameters. Such methodology is referred Gradient Re-parameterization, and named RepOptimizers. For extreme simplicity model structure, focus on VGG-style plain showcase...
We revisit large kernel design in modern convolutional neural networks (CNNs). Inspired by recent advances vision transformers (ViTs), this paper, we demonstrate that using a few kernels instead of stack small could be more powerful paradigm. suggested five guidelines, e.g., applying re-parameterized depth-wise convolutions, to efficient high-performance large-kernel CNNs. Following the propose RepLKNet, pure CNN architecture whose size is as 31x31, contrast commonly used 3x3. RepLKNet...
With the growth of online news over past decade, empirical studies on political discourse and consumption have focused phenomenon filter bubbles echo chambers. Yet recently, scholars revealed limited evidence around impact such phenomenon, leading some to argue that partisan segregation across audiences can- not be fully explained by alone role traditional legacy media may as salient in polarizing public current events. In this work, we expand scope analysis include both more investigating...
Excessive elevation of arterial blood pressure (BP) at high altitude can be detrimental to our health due acute mountain sickness (AMS) or some AMS symptoms. This prospective and observational study aimed elucidate changes induced by exposure high-altitude hypoxia the relationships these with prevalence, severity, sleep quality exercise condition in healthy young men.A was performed 931 male adults exposed 3,700 m (Lhasa) from low (LA, 500 m). Blood measurement symptom questionnaires were LA...
In this work, we examine the linguistic signature of online racial microaggressions (acts) and how it differs from that personal narratives recalling experiences such aggressions (recalls) by Black social media users. We manually curate annotate a corpus acts recalls fromin-the-wild discussions, verify labels with workshop participants. leverage Natural Language Processing (NLP) qualitative analysis on data to classify (RQ1), interpret (RQ2), characterize (RQ3) language underlying in context...
Job interviews play a critical role in shaping one's career, yet practicing interview skills can be challenging, especially without access to human coaches or peers for feedback. Recent advancements large language models (LLMs) present an opportunity enhance the practice experience. Yet, little research has explored effectiveness and user perceptions of such systems benefits challenges using LLMs practice. Furthermore, while prior work recent commercial tools have demonstrated potential AI...
Despite the increasing use of large language models (LLMs) in everyday life among neurodivergent individuals, our knowledge how they engage with and perceive LLMs remains limited. In this study, we investigate individuals interact by qualitatively analyzing topically related discussions from 61 communities on Reddit. Our findings reveal 20 specific LLM cases across five core thematic areas users: emotional well-being, mental health support, interpersonal communication, learning, professional...
To observe the effect of electroacupuncture (EA) bilateral "Jiaji" (EX-B2) on autophagy, apoptosis and phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target rapamycin (mTOR) signaling pathway nucleus pulposus cells in rabbits with intervertebral disk degeneration (IVDD), so as to explore its possible mechanisms underlying improvement IVDD. A total 18 male New Zealand were randomly equally divided into sham operation, model EA groups. The IVDD was established by axial disc...
Background: The aim of this study was to explore the relationship between age and acute mountain sickness (AMS) when subjects are exposed suddenly high altitude. Methods: A total 856 young adult men were recruited. Before after altitude exposure, Athens Insomnia Scale score (AISS) used evaluate subjective sleep quality subjects. AMS assessed using Lake Louise scoring system. Heart rate (HR) arterial oxygen saturation (SaO 2 ) measured. Results: Results showed that, at 500 m, AISS insomnia...
The decreased expression and dysfunction of glucose transporter 4 (GLUT4), the insulin-responsive transporter, are closely related to occurrence insulin resistance (IR).To improve GLUT4 may represent a promising strategy prevent treat IR type 2 diabetes (T2DM).Here, we demonstrate that natural compound tectorigenin (TG) enhances expression, uptake responsiveness via activating AMP-activated protein kinase (AMPK)/myocyte enhancer factor (MEF2) signaling in both normal skeletal muscle cells...
Aim: To determine the prevalence of pulmonary hypertension (PH) and its associated factors among end-stage renal disease (ESRD) patients who underwent maintenance dialysis. Methods: A total 491 received echocardiography examinations artery systolic pressure (PASP) assessments. subgroup 283 were subjected to plasma creatinine (Cr) blood urea nitrogen concentration (BUN) tests, routine electrolyte analysis. First, we compared differences in echocardiographic, Cr BUN, parameters between PH...
Male novel coronavirus disease (COVID-19) patients tend to have poorer clinical outcomes than female patients, while the myocardial injury is strongly associated with COVID-19-related adverse events. Owing a lack of corresponding data, we aimed investigate sex differences in incidence COVID-19 and identify potential underlying mechanisms, which may partly account for bias This retrospective study included 1,157 who were hospitalized Huoshenshan Hospital from 12 March 2020 11 April 2020. Data...
Hypertension is the most common comorbidity in patients with coronavirus disease 2019 (COVID-19) and increases in-hospital mortality. Day-by-day blood pressure (BP) variability (BPV) associated clinical outcomes hypertensive patients. However, little information available on association of BPV COVID-19 hypertension. This study aimed to demonstrate whether day-by-day had prognostic significance these The authors included 702 hypertension from Huoshenshan Hospital (Wuhan, China), who underwent...
Background: A large proportion of populations suffer from acute mountain sickness (AMS) after exposure at high altitude. AMS is closely related with age and gender implying that the sex hormones may play critical roles in AMS. Our observational study aimed to identify association between endogenous testosterone (T), estradiol (E2) Methods: total 113 subjects were recruited 2012. The participants evaluated 500 m (one day) short-term (seven days) high-altitude 3700 m. also completed a case...
We empirically investigate proper pre-training methods to build good visual tokenizers, making Large Language Models (LLMs) powerful Multimodal (MLLMs). In our benchmark, which is curated evaluate MLLMs semantic understanding and fine-grained perception capabilities, we discussed different tokenizers pre-trained with dominant (i.e., DeiT, CLIP, MAE, DINO), observe that: i) Fully/weakly supervised models capture more semantics than self-supervised models, but the gap narrowed by scaling up...
Hyperuricemia has been demonstrated to be a risk factor for cardiovascular diseases. Though the association between uric acid (UA) and arterial stiffness investigated previously in patients with basic diseases, predictive value of baseline UA level not conducted. We aimed at identifying role prospectively. A longitudinal follow-up study routine health check-up population was performed an average 4.8 years. The demographic information, anthropometric parameters, (pulse-wave velocity, PWV),...
End-to-end (E2E) automatic speech recognition models like Recurrent Neural Networks Transducer (RNN-T) are becoming a popular choice for streaming ASR applications voice assistants. While E2E very effective at learning representation of the training data they trained on, their accuracy on unseen domains remains challenging problem. Additionally, these require paired audio and text data, computationally expensive difficult to adapt towards fast evolving nature conversational speech. In this...