A5101904472- Topic Modeling
- Advanced Computational Techniques and Applications
- Sentiment Analysis and Opinion Mining
- Medical Imaging Techniques and Applications
- Industrial Technology and Control Systems
- Guidance and Control Systems
- Advanced Radiotherapy Techniques
- Advanced Text Analysis Techniques
- Military Defense Systems Analysis
- Natural Language Processing Techniques
- Adaptive Control of Nonlinear Systems
- Manufacturing Process and Optimization
- Simulation and Modeling Applications
- Stability and Control of Uncertain Systems
- Advanced Algorithms and Applications
- Power Systems and Technologies
- Distributed and Parallel Computing Systems
- Technology and Security Systems
- Radiomics and Machine Learning in Medical Imaging
- Speech and Audio Processing
- Artificial Intelligence in Healthcare and Education
- Fault Detection and Control Systems
- Chinese history and philosophy
- Advanced Research in Science and Engineering
- Tribology and Lubrication Engineering
The University of Texas Southwestern Medical Center
2016-2025
Northeast Normal University
2008-2024
Yunnan University
2017-2024
Shanghai Electric (China)
2021-2024
Guizhou University
2024
Guizhou University of Finance and Economics
2024
University of Rochester
2024
Hebei GEO University
2020-2023
Soochow University
2023
Sichuan Agricultural University
2014-2023
Objective The primary aim of this research was to address the limitations observed in medical knowledge prevalent large language models (LLMs) such as ChatGPT, by creating a specialized model with enhanced accuracy advice. Methods We achieved adapting and refining meta-AI (LLaMA) using dataset 100,000 patient-doctor dialogues sourced from widely used online consultation platform. These conversations were cleaned anonymized respect privacy concerns. In addition refinement, we incorporated...
The primary aim of this research was to address the limitations observed in medical knowledge prevalent large language models (LLMs) such as ChatGPT, by creating a specialized model with enhanced accuracy advice. We achieved adapting and refining meta-AI (LLaMA) using dataset 100,000 patient-doctor dialogues sourced from widely used online consultation platform. These conversations were cleaned anonymized respect privacy concerns. In addition refinement, we incorporated self-directed...
Compared with conventional word embeddings, sentiment embeddings can distinguish words similar contexts but opposite sentiment. They be used to incorporate information from labeled corpora or lexicons by either end-to-end training refinement. However, these methods present two major limitations. First, traditional approaches provide a fixed representation each ignore the alternation of meaning in different contexts. As result, polarity certain emotional may vary context, will assigned same...
ASVspoof 5 is the fifth edition in a series of challenges which promote study speech spoofing and deepfake attacks as well design detection solutions. We introduce database generated crowdsourced fashion from data collected diverse acoustic conditions (cf. studio-quality for earlier databases) ~2,000 speakers ~100 earlier). The contains with 32 different algorithms, also crowdsourced, optimised to varying degrees using new surrogate models. Among them are mix legacy contemporary...
Abstract Background Unsupervised domain adaptation (UDA) seeks to mitigate the performance degradation of deep neural networks when applied new, unlabeled domains by leveraging knowledge from source domains. In medical image segmentation, prevailing UDA techniques often utilize adversarial learning address shifts for cross‐modality adaptation. Current research on tends adopt increasingly complex models and loss functions, making training process highly intricate less stable/robust....
Applications of deep learning (DL) are essential to realizing an effective adaptive radiotherapy (ART) workflow. Despite the promise demonstrated by DL approaches in several critical ART tasks, there remain unsolved challenges achieve satisfactory generalizability a trained model clinical setting. Foremost among these is difficulty collecting task-specific training dataset with high-quality, consistent annotations for supervised applications. In this study, we propose tailored framework...
Plant essential oils are widely used in food, medicine, cosmetics, and other fields because of their bacteriostatic properties natural sources. However, the range unilateral is limited, compound oil has become an effective way to improve antibacterial oils. In this study, based on analysis Chinese cinnamon bark oregano oil, proportion concentration were optimized designed, activity was studied. The results showed that higher than oil. prepared by a 1:1 ratio with 156.25 ppm excellent against...
We implemented the sentiment system in all five subtasks for English and Spanish. All involve emotion or intensity prediction (regression ordinal classification) emotions determining (multi-labels classification). The useful BiLSTM (Bidirectional Long-Short Term Memory) model with attention mechanism was mainly applied our system. use order to get word information extracted from both directions. used find contribution of each improving scores. Furthermore, based on BiLSTMATT (BiLSTM...
The recent developments of foundation models in computer vision, especially the Segment Anything Model (SAM), allow scalable and domain-agnostic image segmentation to serve as a general-purpose tool. In parallel, field medical has benefited significantly from specialized neural networks like nnUNet, which is trained on domain-specific datasets can automatically configure network tailor specific challenges. To combine advantages models, we present nnSAM, synergistically integrates SAM model...
Personal attributes have been proven to be useful for sentiment analysis. However, previous models of learning attribute-specific language representations are suboptimal because only context- or content-wise injection is adopted. This study proposes a transformer structure with combination both and injections based on well pretrained encoder. For context-wise injection, self-interactive attention implemented by incorporating personal into multi-head attention. the perspective, an...