Abstract The proliferation of deepfake speech poses a significant threat to cybersecurity, from manipulating political speeches and impersonating public figures to spoofing voice biometric systems. The increasing sophistication of adversaries increases the necessity of deploying adaptive detection methods. Moreover, real-world incidents such as fraudulent financial transactions highlight the severity of the problem. Although numerous detectors have been developed, their evaluation remains difficult due to different methodologies and benchmark datasets, making direct comparisons impossible. This study presents a general and detailed framework for evaluating and comparing deepfake speech detectors. We further demonstrate the use of this framework to evaluate 40 state-of-the-art deepfake speech detectors under various conditions and data samples. We objectively compare these methods and identify the key attributes influencing performance the most. We also stress the issue of generalisation, as current detectors struggle to detect previously unseen deepfake speech samples or samples that have been modified. Finally, to strengthen the defence against synthetic audio content, we provide recommendations for improving the robustness of future detectors.

This paper focuses on the problem of detection and prevention of stolen and misused secrets (such as private keys) for authentication toward centralized services. We propose a solution for this problem, based on SmartOTPs, the two-factor authentication scheme against the blockchain, which is intended for smart contract wallets and utilizes one-time passwords (OTPs). We modify SmartOTPs for our purposes and utilize them in the setting of two-and-a-half-factor authentication against a centralized service provider. Out of two and a half factors of our solution, the first factor stands for the private key, and the second and a half factor stands for OTPs and their precursors (a.k.a., pre-images), where OTPs are obtained from the precursors by cryptoaraphically secure hashing. We describe the protocol for bootstrapping our approach as well as the authentication procedure. In the case of stolen creden-tials from the client, we show that our solution enables the user to immediately detect it and proceed to re-initialization with fresh credentials. We utilize blockchain-based identity management and decentralized identities of users to simplify the overhead of the registration process and reinitialization.

Wireless sensor networks usually require certain level of security, however there is lack of sensor nodes with cryptographic hardware build-in. We proved realization of cryptographic platform based on smart card-sensor node connection. Time complexity and power consumption of the platform was measured and compared to conventional software implementation. Our results showed low power demand and high efficiency of public key cryptography running on the platform.