Enhancing Precision in Facial Micro- Expression Recognition with Support Vector Machine in Comparison to ELM Algorithm
T. J. Nagalakshmi, Anji Reddy · 2024
The primary objective of this study is to develop a new method for recognising facial micro-expressions by comparing the accuracy of the SVM and ELM algorithms. The extreme learning machine (N=10) and support vector machine (N=10) used a sample size of 10 and 20 iterations respectively to identify micro expressions with g power of 80%, limit of 0.608, and confidence stretch of 95%. In comparison to extreme learning machines, which have an accuracy of 91.0%, support vector machines will deliver superior exactness (94.8%). The difference in statistical significance, p = 0.608 (p<0.05), According to the free example T -test value, the study's findings are not statistically significant. When it comes to detecting facial micro-expressions, support vector machine algorithms outperform extreme learning machine methods in terms of accuracy.