Can a Robot Catch You Lying? A Machine Learning System to Detect Lies During Interactions

Abstract

Deception is a complex social skill present in human interactions. Many social professions such as teachers, therapists and law enforcement officers leverage on deception detection techniques to support their work activities. Robots with the ability to autonomously detect deception could provide an important aid to human-human and human-robot interactions. The objective of this work is to demonstrate the possibility to develop a lie detection system that could be implemented on robots. To this goal, we focus on human and human robot interaction to understand if there is a difference in the behavior of the participants when lying to a robot or to a human. Participants were shown short movies of robberies and then interrogated by a human and by a humanoid robot “detectives.” According to the instructions, subjects provided veridical responses to half of the question and false replies to the other half. Behavioral variables such as eye movements, time to respond and eloquence were measured during the task, while personality traits were assessed before experiment initiation. Participant's behavior showed strong similarities during the interaction with the human and the humanoid. Moreover, the behavioral features were used to train and test a lie detection algorithm. The results show that the selected behavioral variables are valid markers of deception both in human-human and in human-robot interactions and could be exploited to effectively enable robots to detect lies.

Truth and lies in your eyes: Pupil dilation of White participants in truthful and deceptive responses to White and Black partners

“…The result that race affects pupil dilation as a cue to deception may have implications for the development of more efficient lie detection machines, especially technologies that use behavioral cues associated with deception. For instance, Gonzalez-Billandon et al [76] have recently developed a machine learning system to detect lies that is based on several behavioral cues, including pupil dilation. Our findings suggest that race should be considered in the development of such technologies and that its effects should be further tested considering other relevant behavioral cues to deception.…”

Magic iCub

“…Hence, the measure should not be limited to a specific set of items. is is novel with respect to the state-of-the-art cognitive load assessment methods based on long, tedious and strictly constrained tasks [51], [78], [79] and cumbersome sensing devices. Hence, it represents a step toward those applications where robots could take benefit from evaluating the human partner's internal state and change their behavior accordingly (e.g., by providing a less challenging task).…”

“…Even if the magic trick is based only on the right eye pupil dilation, the features for both eyes are logged on YARP for further analysis. We decided to focus on right-eye features since prior findings on lie detection based on pupillometric features [51] and Tobii documentation [66] reported no significant difference between the two eyes. We also decided to skip the Tobii Pro Glasses 2 eyetracker calibration to not impact the informality of the interaction.…”

“…In the game, iCub has to detect in real-time the player's secret card -the card about which the human is lyingfrom a set of six random cards, during a quick and ecological social interaction. e approach is inspired by previous findings on lie detection in HRI [50], [51] based on cognitive load assessment [52], [53] via pupillometric features [54], [55]. We propose an autonomous end-to-end (E2E) architecture which integrates the cognitive load assessment, the decision making and the robot control, enabling iCub to lead the magic trick with no need of a Wizard of Oz control configuration.…”

Augmented Reality Meets Artificial Intelligence in Robotics: A Systematic Review

“…Recently, the advent of big data and the Internet of Things (IoT), supercomputers, and cheap accessible storage have paved the way for a long-awaited renaissance in artificial intelligence. Currently, research in AI is involved in many domains including robotics ( Le et al, 2018 ; Gonzalez-Billandon et al, 2019 ), natural language processing (NLP) ( Bouaziz et al, 2018 ; Mathews, 2019 ), and expert systems ( Livio and Hodhod, 2018 ; Nicolotti et al, 2019 ). It is becoming ubiquitous in almost every field that requires humans to perform intelligent tasks like detecting fraudulent transactions, diagnosing diseases, and driving cars on crowded streets.…”