A5100652476- Hearing Loss and Rehabilitation
- Advanced Memory and Neural Computing
- Robot Manipulation and Learning
- Speech and Audio Processing
- Advanced Wireless Network Optimization
- Building Energy and Comfort Optimization
- Topic Modeling
- Advanced MIMO Systems Optimization
- Natural Language Processing Techniques
- Acoustic Wave Phenomena Research
- Soft Robotics and Applications
- Urban Heat Island Mitigation
- Hydrocarbon exploration and reservoir analysis
- Web Data Mining and Analysis
- Geochemistry and Geologic Mapping
- Cooperative Communication and Network Coding
- Neural dynamics and brain function
- Advanced Adaptive Filtering Techniques
- Computer Graphics and Visualization Techniques
- Ferroelectric and Negative Capacitance Devices
- Machine Learning in Bioinformatics
- Technology Assessment and Management
- Transportation Systems and Infrastructure
- Enhanced Oil Recovery Techniques
- Mineral Processing and Grinding
China University of Geosciences
2023-2025
DeepMind (United Kingdom)
2024
Zhejiang A & F University
2021-2024
Western Sydney University
2015-2024
Shanghai University
2024
Shanghai University of Traditional Chinese Medicine
2024
Google (United States)
2021-2024
Tongji University
2024
Shanghai East Hospital
2024
Huazhong University of Science and Technology
2024
Abstract Geochemical survey data are a type of spatial big that play an increasingly significant role in mineral exploration. One challenge the era is how to mine geochemical support In this study, based on generative adversarial network framework, we proposed unsupervised spatial–spectrum dual‐branch deep learning method for anomaly identification, namely dual‐DL, which consists branch and spectrum branch. The was constructed using convolutional neural autoencoder, can effectively capture...
Ying Xu, Xu Zhong, Antonio Jimeno Yepes, Jey Han Lau. Proceedings of the 2021 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies. 2021.
Shotgun proteomics protocols are widely used for the identification and/or quantitation of proteins in complex biological samples. Described here is a shotgun protocol that can be to identify protein targets biologically relevant ligands mixtures. The combines quantitative platform with covalent modification strategy, termed Stability Proteins from Rates Oxidation (SPROX), which utilizes denaturant dependence hydrogen peroxide-mediated oxidation methionine side chains assess thermodynamic...
With escalating mortality rates associated with cardiovascular disease, the early detection of arrhythmias assumes ever-increasing significance. This study introduces a novel multimodal network that concurrently classifies electrocardiogram (ECG) and wrist pulse signal (WPS). Both ECG WPS, as human physiological signals, share closely related distributions characteristics, holding potential to accurately reflect underlying conditions. The proposed ICMT-Net utilizes continuous wavelet...
This paper presents a digital implementation of the Cascade Asymmetric Resonators with Fast-Acting Compression (CAR-FAC) cochlear model. The CAR part simulates basilar membrane's (BM) response to sound. FAC models outer hair cell (OHC), inner (IHC), and medial olivocochlear efferent system functions. feeds back by moving poles zeros resonators automatically. We use time-multiplexing pipeline parallelizing techniques implement 70-section, 44.1 kHz real-time CAR-FAC on an Altera Cyclone V...
Unsupervised feature extraction algorithms form one of the most important building blocks in machine learning systems. These are often adapted to event-based domain perform online neuromorphic hardware. However, not designed for purpose, such typically require significant simplification during implementation meet hardware constraints, creating trade offs with performance. Furthermore, conventional generate useful intermediary signals which valuable only context limitations. In this work a...
Phosphorylation is the most important type of protein post-translational modification. Accordingly, reliable identification kinase-mediated phosphorylation has implications for functional annotation phosphorylated substrates and characterization cellular signalling pathways. The local sequence context surrounding potential sites considered to harbour relevant information site prediction models. However, currently there a lack condensed vector representation this contextual information,...
Soft robotic gloves have shown great potential in accelerating the rehabilitation process of individuals with hand pathologies. However, most existing soft assistive devices allow a single degree-of-freedom (DoF) movement for each finger while independent motion joints plays crucial role rehabilitation. Trying to address these challenges, novel fishbone-inspired actuator multi-DoFs is proposed this article first time, best knowledge authors, and preliminary glove developed. With assistance...
Soft robotics have shown tremendous potential in helping patients with hand pathologies perform rehabilitation exercises because of the affordability and inherent softness soft actuators possess. To date, most are fabricated shape deposition manufacturing (SDM) method, which is time-consuming might not be able to withstand high pressure or preserve air tightness after multiple inflationdeflation cycles. In this paper, we present a novel monolithic fabrication method combining advantages...
We present an end-to-end trainable modular event-driven neural architecture that uses local synaptic and threshold adaptation rules to perform transformations between arbitrary spatio-temporal spike patterns. The represents a highly abstracted model of existing Spiking Neural Network (SNN) architectures. proposed Optimized Deep Event-driven network Architecture (ODESA) can simultaneously learn hierarchical features at multiple time scales. ODESA performs online learning without the use error...
Fruit detection and recognition has an important impact on fruit vegetable harvesting, yield prediction growth information monitoring in the automation process of modern agriculture, actual complex environment orchards poses some challenges for accurate detection. In order to achieve green fruits orchard environments, this paper proposes object method based optimized YOLOX_m. First, model extracts features from input image using CSPDarkNet backbone network obtain three effective feature...
We report the results of our experiments in context NEWS 2015 Shared Task on Transliteration.We focus methods combining multiple base systems, and leveraging transliterations from languages.We show error reductions over best system up to 10% when using supplemental transliterations, 20% combination.We also discuss quality shared task datasets.
This paper addresses resource management problem in a heterogeneous network consisting of cellular users and multiple D2D pairs. The share spectrum with uplink under the QoS constraint system pay for interference they cause. We propose two-step scheme to optimize user's transmitting power efficiency network. Firstly, pricing allocation between pair their allocated is formulated as Stackelberg game. Then equilibrium, i.e., optimal price that maximize utilities two players (the base station...
One challenging issue in speaker identification (SID) is to achieve noise-robust performance. Humans can accurately identify speakers, even noisy environments. We leverage our knowledge of the function and anatomy human auditory pathway design SID systems that better performance than conventional approaches. propose a text-dependent system based on real-time cochlear model called cascade asymmetric resonators with fast-acting compression (CARFAC). investigate CARFAC signals corrupted by...
Efficiently increasing the production of clinical-grade mesenchymal stem cells (MSCs) is crucial for clinical applications. Challenges with current planar culture methods include scalability issues, labour intensity, concerns related to cell senescence, and heterogeneous responses. This study aimed establish a large-scale system MSC generation. In addition, comparative analysis biological differences between MSCs cultured under various conditions was conducted.