- Speech Recognition and Synthesis
- Music and Audio Processing
- Speech and Audio Processing
- Thyroid Disorders and Treatments
- Transportation and Mobility Innovations
- Thyroid Cancer Diagnosis and Treatment
- Chinese history and philosophy
- Vehicle Routing Optimization Methods
- Transportation Planning and Optimization
- Natural Language Processing Techniques
- Emotion and Mood Recognition
- Thermochemical Biomass Conversion Processes
- Smart Parking Systems Research
- Traffic control and management
- Phytochemicals and Antioxidant Activities
- RNA modifications and cancer
- Genomics and Chromatin Dynamics
- Single-cell and spatial transcriptomics
- Urban and Freight Transport Logistics
- Topic Modeling
- Chromosomal and Genetic Variations
- Glass properties and applications
- Cancer-related molecular mechanisms research
- Genomic variations and chromosomal abnormalities
- Nanocomposite Films for Food Packaging
Yunnan Institute of Endemic Diseases Control and Prevention
2025
Xinjiang Agricultural University
2025
Harbin Medical University
2021-2024
Chinese Center For Disease Control and Prevention
2022-2024
National Health and Family Planning Commission
2024
Group Sense (China)
2023
Shandong University
2010-2022
Stanford University
2022
Southwest University
2022
Singapore Management University
2016-2021
Molecular characterization of cell types using single-cell transcriptome sequencing is revolutionizing biology and enabling new insights into the physiology human organs. We created a reference atlas comprising nearly 500,000 cells from 24 different tissues organs, many same donor. This enabled molecular more than 400 types, their distribution across tissues, tissue-specific variation in gene expression. Using multiple single donor identification clonal T between mutation rate B cells,...
Abstract Cell-free RNA from liquid biopsies can be analyzed to determine disease tissue of origin. We extend this concept identify cell types origin using the Tabula Sapiens transcriptomic atlas as well individual atlases in combination with Human Protein Atlas consensus dataset. define type signature scores, which allow inference that contribute cell-free for a variety diseases.
Abstract Little is known about the association between iodine nutrition status and bone health. The present study aimed to explore connection status, metabolism parameters, disease (osteopenia osteoporosis). A cross-sectional survey was conducted involving 391, 395, 421 adults from fortification areas (IFA), adequate (IAA), excess (IEA) of China. Iodine parameters BMD were measured. Our results showed that, in IEA, urine concentrations (UIC) serum (SIC) significantly higher than IAA. Ca 2+...
Transformer recently has achieved impressive success in a number of domains, including machine translation, image recognition, and speech recognition. Most the previous work on Keyword Spotting (KWS) is built upon convolutional or recurrent neural networks. In this paper, we explore family architectures for keyword spotting, optimizing trade-off between accuracy efficiency high-speed regime. We also studied effectiveness summarized principles applying key components vision Transformers to...
Effective management of land-side transportation provides the competitive advantage to port terminal operators in improving services and efficient use limited space an urban port. We present a hybrid simulation model that combines traffic-flow modeling discrete-event for planning evaluation traffic conditions number what-if scenarios. design our based on real-world case bulk cargo The problem is interesting due complexity heterogeneous closed-looped internal vehicles external traveling...
Multimodal Sentiment Analysis (MSA) is a popular research topic aimed at utilizing multimodal signals for understanding human emotions. The primary approach to solving this task develop complex fusion techniques. However, the heterogeneity and unaligned nature between modalities pose significant challenges fusion. Additionally, existing methods lack consideration efficiency of modal To tackle these issues, we propose AcFormer, which contains two core ingredients: i) contrastive learning...
In this work, we focus on the robustness of keyword spotting (KWS) at various speech speeds. First, to enable small-footprint KWS, graft a depthwise separable convolution and dilated temporal build our basic model block. Second, make KWS desensitized rate, simple yet effective soft self-attention is proposed operate between different hierarchical features, offering ability be aware speaker's rate dynamically integrate multi-scale features from varying sizes receptive fields. Besides,...
In this paper, we present the solutions to MER-SEMI subchallenge of Multimodal Emotion Recognition Challenge (MER 2023). This focuses on predicting discrete emotions for a small subset unlabeled videos within context semi-supervised learning. Participants are provided with combination labeled and large amounts videos. Our preliminary experiments demonstrate that task is primarily driven by video audio modalities, while text modality plays relatively weaker role in emotion prediction. To...
In this paper, we present the solutions to MER-MULTI and MER-NOISE sub-challenges of Multimodal Emotion Recognition Challenge (MER 2023). For tasks MER-NOISE, participants are required recognize both discrete dimensional emotions. Particularly, in test videos corrupted with noise, necessitating consideration modality robustness. Our empirical findings indicate that different modalities contribute differently tasks, a significant impact from audio visual modalities, while text plays weaker...
Torreya grandis is a unique tree species in China. Although full use has been made of the timber, processing and utilization bark not effective. In order to explore new way utilize grandis, powder T. was prepared analyzed qualitatively quantitatively. Differential scanning calorimetry (TG) pyrolysis gas chromatography-mass spectrometry (PY-GC/MS) revealed many bioactive components such as acetic acid, 2-methoxy-4-vinyl phenol, D-mannose, furfural. These substances have potential broad...
Grapheme-to-phoneme (G2P) conversion is an indispensable part of the Chinese Mandarin text-to-speech (TTS) system, and core G2P to solve problem polyphone disambiguation, which pick up correct pronunciation for several candidates a polyphonic character.In this paper, we propose BERT model predict pronunciations characters.Firstly, create 741 new monophonic characters from 354 source by pronunciation.Then get extending pre-trained with adding corresponding embedding layer tokens, initialized...
Abstract Iodine excess (IE) can cause thyroid dysfunction, diseases adversely affect cardiovascular function. Accordingly, this study was to explore the direct and indirect effects of IE on endothelial Nthy‐ori 3‐1 HUVECs cells were treated with potassium iodide (KI). CCK‐8, LDH leakage, Elisa, RT‐PCR Western blotting used detect relevant indicators. Results showed that a certain level KI directly indirectly reduce viability increase cytotoxicity. decreased expression ET‐1 VWF in HUVECs,...
Abstract Iodine is an essential nutrient that may change the occurrence of autoimmune thyroiditis (AIT). Apoptosis and DNA methylation participate in pathogenesis destructive mechanism AIT. We detected expression mRNA intrinsic apoptosis-associated genes ( YWHAG , ING4 BRSK2 GJA1 ) to identify potential interactions between levels these different iodine. 176 adult patients with AIT Shandong Province, China, were included. The MethylTarget TM assay was used verify methylation. PCR detect...