Explainable Label-flipping Attacks on Human Emotion Assessment System

Abstract--This paper's main goal is to provide an attacker's The dataset was compiled to look at potential applications as current or past employees, and have inside knowledge of the of brainwave signals for spotting insider threats in the business's security protocols, customer information, and workplace. The Emotiv Insight 5 channels were the tool used computer systems. To assess human emotions and acts, speech to collect the data. Information from 17 people who gave their [1] and facial expression [2] data were employed traditionally. On the other hand, EEG The four risk categories--High-Risk, Medium-Risk, Low-signals [3] have been utilized in recent years to assess a Risk, and Normal--found in the risk matrix were used to person's emotional state to prevent potential industrial insider classify each signal for a captured image, and each signal was assaults because people cannot conceal or manipulate their then given the appropriate label.