A5101708658- Privacy-Preserving Technologies in Data
- Gaze Tracking and Assistive Technology
- User Authentication and Security Systems
- Internet Traffic Analysis and Secure E-voting
- Biometric Identification and Security
- Cryptography and Data Security
- Generative Adversarial Networks and Image Synthesis
- Wireless Body Area Networks
- Smart Grid Security and Resilience
- Advanced Measurement and Detection Methods
- Advanced Vision and Imaging
- Anomaly Detection Techniques and Applications
- Water Quality Monitoring Technologies
- Adversarial Robustness in Machine Learning
- Context-Aware Activity Recognition Systems
- Recommender Systems and Techniques
- Advanced Image and Video Retrieval Techniques
- Advanced Data Storage Technologies
- Cloud Data Security Solutions
- Machine Learning in Healthcare
- Digital Media Forensic Detection
- Video Surveillance and Tracking Methods
- Non-Invasive Vital Sign Monitoring
- Music and Audio Processing
- Mobile Crowdsensing and Crowdsourcing
University of Helsinki
2022-2025
University of California, Berkeley
2025
Southwest Minzu University
2024
Tianjin University
2019-2020
Beihang University
2020
Virtual Reality Medical Center
2018
State Key Laboratory of Virtual Reality Technology and Systems
2018
The Internet of Things (IoT) has revolutionized the connectivity diverse sensing devices, generating an enormous volume data. However, applying machine learning algorithms to devices presents substantial challenges due resource constraints and privacy concerns. Federated (FL) emerges as a promising solution allowing for training models in distributed manner while preserving data on client devices. We contribute SAFI , semi-asynchronous FL approach based clustering achieve novel in-cluster...
Foundation models (FMs) have achieved state-of-the-art performance across various domains, benefiting from their vast number of parameters and the extensive amount publicly available training data. However, real-world deployments reveal challenges such as system heterogeneity, where not all devices can handle complexity FMs, emerging privacy concerns that limit availability public To address these challenges, we propose HeLoRA, a novel approach combining low-rank adaptation (LoRA) with...
Screen lock is a critical security feature for smart-phones to prevent unauthorized access. Although various screen unlocking technologies including fingerprint and facial recognition have been widely adopted, they still some limitations. For example, fingerprints can be stolen by special material stickers systems cheated 3D-printed head models. In this paper, we propose EmgAuth, novel electromyography(EMG)-based smartphone system based on the Siamese network. EmgAuth leverages Myo armband...
Privacy protection in recommendation systems is gaining increasing attention, for which federated learning has emerged as a promising solution. Current grapple with high communication overhead due to sharing dense global embeddings, and also poorly reflect user preferences data heterogeneity. To overcome these challenges, we propose two-stage Federated Low-rank Coordinated Adaptation (FedLoCA) framework decouple client-specific knowledge into low-rank significantly reduces while enhancing...
Recent years have witnessed an increasing interest in online video affective content analysis, since having a better understanding of the emotions videos can facilitate many applications including retrieval and classification. Research computing requires ground truth data for training evaluation. The live commentary (also known as 'danmaku', 'barrage', or 'bullet comment') is quite popular recent years, but few researchers paid attention to information analysis videos. In this paper, we...
Internet of Things (IoT) interconnects a massive amount devices, generating heterogeneous data with diverse characteristics. IoT emerges as vital asset for data-intensive applications, such healthcare, smart city and predictive maintenance, harnessing the vast volume to its maximum advantage. These applications leverage different Artificial Intelligence (AI) algorithms discover new insights. While machine learning effectively uncovers implicit patterns through model training, centralizing...
Among various tools invented to help improve people's oral health, water flossers can achieve better performance than traditional and electronic toothbrushes, are less harmful dental floss, especially for those with orthodontic teeth or tooth implant surgeries. However, the available in market serve no monitoring recording functions that consumers clean their a more efficient way. To capture users' motions, this study develops novel smart flosser, installing an Inertial Measurement Unit...
Screen lock is a critical security feature for smartphones to prevent unauthorized access. Although various screen unlocking technologies, including fingerprint and facial recognition, have been widely adopted, they still some limitations. For example, fingerprints can be stolen by special material stickers recognition systems cheated 3D-printed head models. In this paper, we propose EmgAuth, novel electromyography(EMG)-based smartphone system based on the Siamese network. EmgAuth enables...
Unsupervised learning with Generative Adversarial Networks (GANs) has made great achievements these years. Generally, traditional GANs model improve the generation effect by proposing new loss functions or adjusting network structures. However, category information of training data is rarely used in and discriminative power GANs' discriminator limited. To overcome such a problem, we propose an improved based on Category Information (CIGAN), which applies hash center to speed CIGAN model. We...
Despite extensive research into data heterogeneity in federated learning (FL), system remains a significant yet often overlooked challenge. Traditional FL approaches typically assume homogeneous hardware resources across clients, implying that clients can train global model within comparable time. However, practical systems, have heterogeneous resources, which impacts their capacity for training tasks. This discrepancy highlights the significance of exploring model-heterogeneous FL, paradigm...
In the tracking-by-detection scheme of multiple object tracking (MOT), data association process in which existing and new detections are matched over time is very important. A framework proposed to solve problem MOT scenarios where there potential target interactions occlusions crowded environments. This consists an input layer layer. The end-to-end feature-map extraction model that incorporates a simplified Siamese convolutional neural network, effectively distinguishes similar objects...
Screen lock is a critical security feature for smartphones to prevent unauthorized access. Although various screen unlocking technologies, including fingerprint and facial recognition, have been widely adopted, they still some limitations. For example, fingerprints can be stolen by special material stickers recognition systems cheated 3D-printed head models. In this paper, we propose EmgAuth, novel electromyography(EMG)-based smartphone system based on the Siamese network. EmgAuth enables...