Deception detection using multimodal fusion approaches

Deep learning driven multimodal fusion for automated deception detection

2017 IEEE Symposium Series on Computational Intelligence (SSCI) ◽

10.1109/ssci.2017.8285382 ◽

2017 ◽

Cited By ~ 13

Author(s):

Mandar Gogate ◽

Ahsan Adeel ◽

Amir Hussain

Keyword(s):

Deep Learning ◽

Deception Detection ◽

Multimodal Fusion

Download Full-text

Response-Specific Scalp Distributions in Deception Detection and ERP Correlates of Psychopathic Personality Traits

Journal of Psychophysiology ◽

10.1027/0269-8803.18.1.13 ◽

2004 ◽

Vol 18 (1) ◽

pp. 13-26 ◽

Cited By ~ 13

Author(s):

Antoinette R. Miller ◽

J. Peter Rosenfeld

Keyword(s):

Deception Detection ◽

Event Related Potential ◽

Psychopathic Personality ◽

P300 Amplitude ◽

Group Differences ◽

Post Hoc Analysis ◽

P300 Component ◽

Psychopathic Personality Inventory ◽

Post Hoc ◽

First Time

Abstract University students were screened using items from the Psychopathic Personality Inventory and divided into high (n = 13) and low (n = 11) Psychopathic Personality Trait (PPT) groups. The P300 component of the event-related potential (ERP) was recorded as each group completed a two-block autobiographical oddball task, responding honestly during the first (Phone) block, in which oddball items were participants' home phone numbers, and then feigning amnesia in response to approximately 50% of items in the second (Birthday) block in which oddball items were participants' birthdates. Bootstrapping of peak-to-peak amplitudes correctly identified 100% of low PPT and 92% of high PPT participants as having intact recognition. Both groups demonstrated malingering-related P300 amplitude reduction. For the first time, P300 amplitude and topography differences were observed between honest and deceptive responses to Birthday items. No main between-group P300 effects resulted. Post-hoc analysis revealed between-group differences in a frontally located post-P300 component. Honest responses were associated with late frontal amplitudes larger than deceptive responses at frontal sites in the low PPT group only.

Download Full-text

Spotting a liar by looking for microexpressions: Challenges and Opportunities

10.31234/osf.io/72yrq ◽

2019 ◽

Author(s):

Xunbing Shen

Keyword(s):

Empirical Evidence ◽

Deception Detection ◽

Empirical Studies ◽

High Stake ◽

Challenges And Opportunities ◽

Cues To Deception ◽

The Relationship

Microexpressions do exist, and they are regarded as valid cues to deception by many researchers, furthermore, there is a lot of empirical evidence which substantiates this claim. However, some researchers don’t think the microexpression can be a way to catch a liar. The author elucidates the theories predicting that looking for microexpressions can be a way to catch a liar, and notes that some data can support for the utilization of microexpressions as a good way to detect deception. In addition, the author thinks that the mixed results in the area of investigating microexpressions and deception detection may be moderated by the stake. More empirical studies which employ high-stake lies to explore the relationship between microexpressions and deception detection are needed.

Download Full-text