A5101515468- Privacy-Preserving Technologies in Data
- Adversarial Robustness in Machine Learning
- Recommender Systems and Techniques
- Anomaly Detection Techniques and Applications
- Neural Networks and Applications
- Advanced Graph Neural Networks
- Sparse and Compressive Sensing Techniques
- Caching and Content Delivery
- Data Quality and Management
- Advanced SAR Imaging Techniques
- Advanced Malware Detection Techniques
- Cytomegalovirus and herpesvirus research
- Tensor decomposition and applications
- Natural Language Processing Techniques
- Advanced Neuroimaging Techniques and Applications
- Peer-to-Peer Network Technologies
- Dementia and Cognitive Impairment Research
- Advanced Image Processing Techniques
- Handwritten Text Recognition Techniques
- Image Processing Techniques and Applications
- Ultrasonics and Acoustic Wave Propagation
- Neurological Complications and Syndromes
- Advanced Vision and Imaging
- Mathematics, Computing, and Information Processing
Zhejiang Lab
2024
National Computer Network Emergency Response Technical Team/Coordination Center of Chinar
2023
Shenzhen Center for Disease Control and Prevention
2022
Guilin University of Electronic Technology
2020-2022
State Key Laboratory of Remote Sensing Science
2020
North Carolina State University
2016
Federated learning is a collaborative machine framework where multiple clients jointly train global model. To mitigate communication overhead, it common to select subset of for participation in each training round. However, existing client selection strategies often rely on fixed number throughout all rounds, which may not be the optimal choice balancing efficiency and model performance. Moreover, these approaches typically evaluate solely based their performances one single round,...
Alzheimer's disease (AD) is a progressive and irreversible neurodegenerative disorder that has recently seen serious increase in the number of affected subjects. In last decade, neuroimaging been shown to be useful tool understand AD its prodromal stage, amnestic mild cognitive impairment (MCI). The majority AD/MCI studies have focused on diagnosis, by formulating problem as classification with binary outcome or healthy controls. There emerged associate image scans continuous clinical scores...
In this article, a supervised nonlinear dictionary learning (DL) method, called multiscale kernel DL (MSK-DL), is proposed for target recognition in synthetic aperture radar (SAR) images. We use Frost filters with different parameters to extract an SAR image's features data augmentation and noise suppression. order reduce the computation cost, dimension of each scale feature reduced by principal component analysis (PCA). Instead widely used linear DL, we learn multiple dictionaries capture...
Recently, deep learning has made significant inroads into the Internet of Things due to its great potential for processing big data. Backdoor attacks, which try influence model prediction on specific inputs, have become a serious threat neural network models. However, because poisoned data used plant backdoor victim typically follows fixed pattern, most existing attacks can be readily prevented by common defense. In this paper, we leverage natural behavior and present stealthy attack image...
Retrieval-Augmented Generation (RAG) has emerged as a key paradigm for enhancing large language models (LLMs) by incorporating external knowledge. However, current RAG methods face two limitations: (1) they only cover limited scenarios. (2) They suffer from task diversity due to the lack of general dataset. To address these limitations, we propose RAG-Instruct, method synthesizing diverse and high-quality instruction data based on any source corpus. Our approach leverages five paradigms,...
Training a machine learning model with data following meaningful order, i.e., from easy to hard, has been proven be effective in accelerating the training process and achieving better performance. The key enabling technique is curriculum (CL), which seen great success deployed areas like image text classification. Yet, how CL affects privacy of unclear. Given that changes way memorizes data, its influence on needs thoroughly evaluated. To fill this knowledge gap, we perform first study...
CNNs facilitate the significant process of single-image super-resolution (SISR). However, most existing CNN-based models suffer from numerous parameters and exceeding deeper structures. Moreover, these relying on deep features commonly ignore hints low-level features, resulting in poor performance. In this paper, we demonstrate an effective network named CASR, which addresses problems by extracting Head Module via strategies based depth-wise separable convolution includes a cascading...