Micro-Expression Recognition for Lie Detection Using Image Processing

Authors- Dr. Pankaj Malik, Akshit Harsola, Somya Sharma, Srashti chouhan, Ananya Subramanya Rao, Akanksha Bhadouriya

Abstract-Micro-expressions are involuntary facial expressions that reveal a person’s true emotions, often occurring in high-stakes situations such as interrogations and security screenings. Detecting these subtle expressions is crucial for lie detection, as they can indicate concealed emotions. This research proposes a novel micro-expression recognition system using image processing techniques to enhance the accuracy of deception detection. The methodology involves facial feature extraction using Local Binary Patterns (LBP) and Optical Flow analysis, followed by classification using a Convolutional Neural Network (CNN). Experiments were conducted on benchmark micro-expression datasets, including CASME II and SAMM, to evaluate system performance. The proposed model achieved an accuracy of 89.4%, outperforming traditional handcrafted feature-based methods. The results demonstrate that deep learning-based image processing techniques can significantly enhance the recognition of micro-expressions for lie detection applications. The findings suggest potential applications in forensic investigations, security protocols, and psychological analysis. Future work will focus on real-time implementation and improving recognition across diverse demographic groups.

DOI: /10.61463/ijset.vol.13.issue2.272