Micro-expression recognition: an updated review of current trends, challenges and solutions

Abstract

Micro-expression (ME) recognition has attracted numerous interests within the computer vision circle in different contexts particularly, localization, magnification, and recognition. Challenges in these areas remain relevant due to the nature of ME’s split-second transition with minute intensity levels. In this paper, a comprehensive state-of-the-art analysis of ME recognition and detection challenges are provided. Contemporary solutions are categorized into low-level, mid-level, and high-level solutions with a review of their characteristics and performances. This paper also provides possible extensions to basic methods, highlight, and predict emerging trends. A thorough analysis of mainstream ME datasets is also provided by elucidating each of their advantages and limitations. This survey gives readers an understanding of ME recognition and an appreciation of future research direction in ME recognition systems.