A5101735569- COVID-19 diagnosis using AI
- Phonocardiography and Auscultation Techniques
- Music and Audio Processing
- Respiratory and Cough-Related Research
- Image Processing and 3D Reconstruction
- Geochemistry and Elemental Analysis
- Geochemistry and Geologic Mapping
- Coastal wetland ecosystem dynamics
- Speech and Audio Processing
- Soil Geostatistics and Mapping
- Single-cell and spatial transcriptomics
- Water Quality and Resources Studies
- Emotion and Mood Recognition
- Marine animal studies overview
- Quality and Safety in Healthcare
- Wood and Agarwood Research
- Statistics Education and Methodologies
- Geological formations and processes
- Advanced Fluorescence Microscopy Techniques
- Gene expression and cancer classification
- Speech Recognition and Synthesis
- Privacy-Preserving Technologies in Data
- Cultural Heritage Materials Analysis
- Molecular Biology Techniques and Applications
- Species Distribution and Climate Change
Imperial College London
1998-2024
Jilin Medical University
2024
Jilin University
2024
National Dong Hwa University
2004
DANCE is the first standard, generic, and extensible benchmark platform for accessing evaluating computational methods across spectrum of datasets numerous single-cell analysis tasks. Currently, supports 3 modules 8 popular tasks with 32 state-of-art on 21 datasets. People can easily reproduce results supported algorithms major via minimal efforts, such as using only one command line. In addition, provides an ecosystem deep learning architectures tools researchers to facilitate their own...
Since the COronaVIrus Disease 2019 (COVID-19) outbreak, developing a digital diagnostic tool to detect COVID-19 from respiratory sounds with computer audition has become an essential topic due its advantages of being swift, low-cost, and eco-friendly. However, prior studies mainly focused on small-scale datasets. To build robust model, large-scale multi-sound FluSense dataset is utilised help cough in this study. Due gap between COVID-19-related datasets consisting only, transfer learning...
Early diagnosis of cardiovascular diseases is a crucial task in medical practice. With the application computer audition healthcare field, artificial intelligence (AI) has been applied to clinical non-invasive intelligent auscultation heart sounds provide rapid and effective pre-screening. However, AI models generally require large amounts data which may cause privacy issues. Unfortunately, it difficult collect from single centre.
The coronavirus disease 2019 (COVID-19) has caused a crisis worldwide. Amounts of efforts have been made to prevent and control COVID-19's transmission, from early screenings vaccinations treatments. Recently, due the spring up many automatic recognition applications based on machine listening techniques, it would be fast cheap detect COVID-19 recordings cough, key symptom COVID-19. To date, knowledge acoustic characteristics cough sounds is limited but essential for structuring effective...
Cardiovascular diseases (CVDs) have been ranked as the leading cause for deaths. The early diagnosis of CVDs is a crucial task in medical practice. A plethora efforts were given to automated auscultation heart sound, which leverages power computer audition develop cheap, non-invasive method that can be used at any time and anywhere measuring status heart. Nevertheless, previous works ignore an important factor, namely, privacy user data. On one hand, learnt models are always hungry bigger...
Abstract Potteries, one of the tools widely used by early humans, encapsulates rich historical information. Deep neural networks have been applied to analyzing pottery digital images, bypassing need for intricate handcrafted features. However, existing models focus solely on shape comparison, neglecting analysis their evolution across different periods. In this work, we propose a method based deep learning assist experts in identifying evolutionary patterns given type within specified...
Speech emotion recognition (SER) has been a popular research topic in human-computer interaction (HCI). As edge devices are rapidly springing up, applying SER to is promising for huge number of HCI applications. Although deep learning investigated improve the performance by training complex models, memory space and computational capability represents constraint embedding models. We propose neural structured (NSL) framework through building synthesized graphs. An model trained on source...
Due to the COronaVIrus Disease 2019 (COVID-19) pandemic, early screening of COVID-19 is essential prevent its transmission. Detecting with computer audition techniques has in recent studies shown potential achieve a fast, cheap, and ecologically friendly diagnosis. Respiratory sounds speech may contain rich complementary information about clinical conditions. Therefore, we propose training three deep neural networks on types (breathing/counting/vowel) assembling these models improve...
Heart sound auscultation is an effective method for early-stage diagnosis of heart disease. The application deep neural networks gaining increasing attention in automated classification. This paper proposes Convolutional Neural Networks (CNNs) to classify normal/abnormal sounds, which takes two-dimensional Mel-scale features as input, including Mel frequency cepstral coefficients (MFCCs) and the Log spectrum. We employ two weighted loss functions during training mitigate class imbalance...
Abstract Purpose The coronavirus disease 2019 (COVID-19) has caused a crisis worldwide. Amounts of efforts have been made to prevent and control COVID-19’s transmission, from early screenings vaccinations treatments. Recently, due the spring up many automatic recognition applications based on machine listening techniques, it would be fast cheap detect COVID-19 recordings cough, key symptom COVID-19. To date, knowledge acoustic characteristics cough sounds is limited, but essential for...
Supervised fine-tuning (SFT) is crucial for aligning Large Language Models (LLMs) with human instructions. The primary goal during SFT to select a small yet representative subset of training data from the larger pool, such that this achieves results comparable or even exceeding those obtained using entire dataset. However, most existing selection techniques are designed small-scale pools, which fail meet demands real-world scenarios. In paper, we replicated several self-scoring methods do...
Ubiquitous sensing has been widely applied in smart healthcare, providing an opportunity for intelligent heart sound auscultation. However, devices contain sensitive information, raising user privacy concerns. To this end, federated learning (FL) adopted as effective solution, enabling decentralised without data sharing, thus preserving the Internet of Health Things (IoHT). Nevertheless, traditional FL requires same architectural models to be trained across local clients and global servers,...
Abstract Imaging-based spatial transcriptomics (iST), such as MERFISH, CosMx SMI, and Xenium, quantify gene expression level across cells in space, but more importantly, they directly reveal the subcellular distribution of RNA transcripts at single-molecule resolution. The localization molecules plays a crucial role compartmentalization-dependent regulation genes within individual cells. Understanding intracellular for particular cell type thus not only improves characterization identity...
Coughs sounds have shown promising as a potential marker for distinguishing COVID individuals from non-COVID ones. In this paper, we propose an attention-based ensemble learning approach to learn complementary representations cough samples. Unlike most traditional schemes such mere maxing or averaging, the proposed fairly considers contribution of representation generated by each single model. The attention mechanism is further investigated at feature level and decision level. Evaluated on...
Abstract Current sediment quality guidelines generally adopt a tiered approach in order to assess more cost‐effectively. The uncertainties involved the of an integrative assessment, however have not been quantified resulting risk committing type I error or II at final confirmatory stage. This study develops statistical criteria and for chemistry component assessment quality. At tier 1 screening stage, historical data initial survey is required determine minimum sample numbers that will be...
Among the 17 Sustainable Development Goals (SDGs) proposed within 2030 Agenda and adopted by all United Nations member states, 13th SDG is a call for action to combat climate change. Moreover, SDGs 14 15 claim protection conservation of life below water on land, respectively. In this work, we provide literature-founded overview application areas, in which computer audition – powerful but context so far hardly considered technology, combining audio signal processing machine intelligence...