Tags → #deep neural networks

Vulnerabilities
Researcher Proposes Watermarking Method for Deep Neural Networks
Nov 22, 2024 Ikuya Morikawa from Fujitsu Limited has published a paper proposing a new defense mechanism called bounding-box watermarking to protect deep neural networks from model extraction attacks.
Vulnerabilities
New Framework Enables Extraction of Deep Neural Networks
Nov 19, 2024 A recent research paper presents a new framework for extracting embedded Deep Neural Networks using side-channel attacks, highlighting significant vulnerabilities in DNNs and raising concerns about the security of models deployed on edge devices.
Vulnerabilities
Research Paper Proposes New Method to Combat Backdoor Attacks
Nov 15, 2024 A research paper from The Chinese University of Hong Kong, Shenzhen, introduces a novel method called Distance-Driven Detoxification (D3) to enhance defenses against backdoor attacks in machine learning models.
Tools
New Framework Developed for Secure Medical Data Sharing
Nov 12, 2024 A new framework for privacy-preserving and secure data sharing in medical image analysis has been developed, integrating advanced encryption techniques and the Vision Transformer model to enhance performance and security while complying with regulations like HIPAA and GDPR.