Abstract

In the detection of suspicious activity using image analysis, standard Convolutional Neural Networks (CNNs) face obstacles such as low accuracy and high computational time. However to improve the CNN model, optimisation strategies like the Spider Wasp Optimiser (SWO) have been explored. In this paper, the spider intelligent optimizer was integrated into a standard CNN model and the resultant optimized convolutional model was trained and tested with 5000 face images acquired from Ladoke Akintola University of Technology students, Faculty of Computing and Informatics, and applied a pre-processing workflow including resizing cropping and grayscale transformation. The optimised model was implemented in MATLAB R20023a. The performance of the formulated model was measured using sensitivity, specificity, precision, accuracy, false positive rate, computation time, and compared against existing standard CNN approach. The results revealed that the SWO-CNN model showed notable performance with a sensitivity of 98.12%, specificity of 97.53%, precision of 98.12%, accuracy of 97.87%, F1-score of 98.12%, and a reduced FPR of 2.47%, taking 71.64 seconds to execute. In contrast, the standard CNN model achieved a sensitivity of 96.71%, specificity of 95.52%, precision of 96.60%, accuracy of 96.20%, and F1-score of 96.66%, with a false positive rate of 4.48% and computational time of 96.04 seconds. This result show that the SWO-CNN offers a superior performance over existing standard CNN.