1 AI-Powered Facial Expression Analysis In Real Time: Unveiling Emotions
D Aishwarya, P. V. K. Sandeep, Penmatsa V P S Akhilesh Varma · 2024
This study introduces a novel facial deconstruction method for face emotion identification. After identifying facial landmarks with the IntraFace algorithm, seven regions of interest (ROI) are extracted, representing the principal featuresof the face: the mouth, nose, an area between the eyebrows, right and left eyes, and eyebrows. Next, a variety of local descriptors, such as Local Binary Patterns, Completed Local Binary Patterns. Local Ternary Patterns, and Dynamic Local Ternary Patterns, are employed to capture distinctive features. Inputting the feature vector which represents a face image into a multi-class support vector machine completes the recognition process. The suggested method outperforms previous facial deconstruction techniques, according on experimental findings on two publicly available datasets.