Toward Stealthy Backdoor Attacks Against Speech Recognition via Elements of Sound

Deep neural networks (DNNs) have been widely and successfully adopted and deployed in various applications of speech recognition. Recently, a few works revealed that these models are vulnerable to backdoor attacks, where the adversaries can implant malicious prediction behaviors into victim models by poisoning their training process. In this...

Inactivation of Escherichia coli and Staphylococcus aureus by using a UVC-LED module with a multi-wavelength setting

UVC-LED is known as a deep ultraviolet LED. The application development and disinfection efficiency of UVC-LED modules are important problems encountered when UVC-LED products are rushed into commercialization. In this article, a specific disinfection experiment with a UVC-LED module was combined to analyze the disinfection efficiency. UVC-LEDs...

Microstructure and hardness of SAC305 and SAC305-0.3Ni solder on Cu, high temperature treated Cu, and graphene-coated Cu substrates

In this study, SAC305 and SAC305-0.3Ni solder balls were soldered onto Cu, high temperature treated Cu (H-Cu) and graphene coated Cu (G-Cu) substrates, respectively. The microstructure, the interfacial reaction, and the hardness of the solder joints were investigated. The interfacial intermetallic compound (IMC) is Cu₆Sn<sub>5</sub...