Micro-expression and Masked Expression Classification Using Neural Networks

Abstract

Micro-expressions are short, involuntary facial expressions unnoticeably occurring for a fraction of a second. They are highly valuable in various real-life situations since the facial muscles associated with micro-expressions cannot be consciously controlled. They can also aid psychiatrists in diagnosing patients or police officers in detecting the true emotions of suspects. In contrast, masked expressions are intended to hide genuine emotions, and these occur when someone uses a falsified expression to conceal their true feeling. In this study, a custom Micro Mask Emotion Dataset-Face Speech Text was developed for both micro-expressions and masked expressions using questionnaires and images as stimuli. The study participants were 18–25 years old and did not wear any face mask. The dataset was compiled and evaluated using various neural network models. The results indicate that females are more expressive and the questionnaire-induced dataset showed better prediction ability than the image-induced dataset on a similar set of neural network models.

Received: 13 May 2025| Revised: 13 August 2025 | Accepted: 2 September 2025

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

Data are available on request from the corresponding author upon reasonable request.

Author Contribution Statement

Karan Suresh: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft. Shamyuktha Ramachandran Sugumaran: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft. Suja Palaniswamy: Conceptualization, Data curation, Writing – original draft, Writing – review & editing, Supervision, Project administration. Manju Priya Arthanarisamy Ramaswamy: Writing – review & editing, Visualization.