A5064468334- Adversarial Robustness in Machine Learning
- Advanced Malware Detection Techniques
- Advanced Neural Network Applications
- EEG and Brain-Computer Interfaces
- Emotion and Mood Recognition
- Privacy-Preserving Technologies in Data
- Fault Detection and Control Systems
- Security and Verification in Computing
- Domain Adaptation and Few-Shot Learning
- Cryptography and Data Security
- Neural Networks and Applications
- Gaze Tracking and Assistive Technology
- Network Security and Intrusion Detection
- Spectroscopy and Chemometric Analyses
- Anomaly Detection Techniques and Applications
Donghua University
2022-2024
As a pioneering distributed learning framework, federated (FL) has gained widespread adoption. It operates collaboratively among participants, with communication limited to sharing model parameters between the server and participants. However, FL is also more susceptible active attacks from malicious insiders. Poisoned updates submitted by attackers can degrade performance of global model. Previous research only considered using naive data clients for backdoor poisoning, therefore achieved...
Transfer-based attacks employ the proxy model to craft adversarial examples against target model, which has seen significant advancements in black-box attacks. Conversely, defense strategies have been devised mitigate such Unfortunately, current attack methods mainly rely on neural network training techniques, as input transformation and gradient regularization, without harnessing mechanisms enhance themselves. In light of this, we propose a novel framework transfer-based with hypothetical...
Electroencephalography (EEG), as a physiological cue, is more objective and reliable in identifying emotions than non-physiological cues. Previous methods only consider one or two relationships among frequency, time spatial domain features of EEG signals, the designed models may still be relatively large terms parameters. Meanwhile, training process previous networks troublesome during algorithm optimization. To address these challenges, we design simple efficient feature preprocessing...