MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641091857 A) filed by Psg College Of Arts & Science on July 29, 2026, for Smart Lie-Stress Differentiation Analyzer (slsda).
Inventors include Balkis Fathima A; Dr. Cyril Allen Jonathan W; Girija N; Banupriya A; Elakkiya M; and Mohammed Faias O.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to an intelligent system of psychological interrogation that could differentiate between a real stress response and forced falsification by multi-modal analysis of physiological and behavioural factors. Lie detection apparatuses based on conventional methods mainly use the polygraph, which can misread anxiety, fear or emotional upset as lying, leading to poor accuracy and reliability. This work overcomes the drawbacks by gathering all the required data from the pulse, pupil, voice, and respiration variability, facial expression microexpressions, cognitive response timing, and the context into a single Artificial Intelligence (Al) platform, which is called Smart Lie-Stress Differentiation Analyzer (SLSDA). A novel Adaptive Cognitive-Emotional Fusion Engine (ACEF) dynamically is able to disassociate stress-related physiological responses from deception-specific behavioural indicators by means of machine learning models learned from multimodal psychophysiological data. The system is continually calculating a Truth Confidence Index (TCI) and Stress Authenticity Score (SAS) and adjusting the interrogation technique according to the individual baseline emotional trait. The invention also provides adaptive baseline calibration, contextual questioning, and confidence-based decision fusion to greatly minimize 'false positives' from anxiety disorders, trauma or high-pressure environments, which eliminates the need for monotonous questioning and complicated algorithms. The invention may be used in border security, military intelligence, behavioural research, clinical research, security screening, psychological evaluation and forensic investigations and for providing explainable AI-based decisions instead of guilt determination.
Disclaimer: Curated by HT Syndication.
