Performance Analysis of Different Optimization Algorithms for Multi-Class Object Detection

Object recognition is a significant approach employed for recognizing suitable objects from the image. Various improvements, particularly in computer vision, are probable to diagnose highly difficult tasks with assistance of local feature detection methodologies. Detecting multi-class quite challenging, and many existing researches have worked enhance overall accuracy. But because certain limitations like higher network loss, degraded training ability, improper consideration features, less convergent so on. The proposed research introduced hybrid convolutional neural (H-CNN) overcome these drawbacks. collected input images pre-processed initially through Gaussian filtering eradicate noise image quality. Followed by pre-processing, present localized using Grid Guided Localization (GGL). effective features extracted AlexNet model. Different classified replacing concluding softmax layer Support Vector Regression (SVR) losses model optimized Improved Grey Wolf (IGW) optimization procedure. performances analyzed PYTHON. datasets employed, including MIT-67, PASCAL VOC2010, Microsoft (MS)-COCO MSRC. varying loss algorithms improved Particle Swarm Optimization (IPSO), Genetic Algorithm (IGA), dragon fly algorithm (IDFA), simulated annealing (ISAA) bacterial foraging (IBFA), choose best algorithm. accuracy outcomes attained as VOC2010 (95.04%), MIT-67 dataset (96.02%), MSRC (97.37%), MS COCO (94.53%), respectively.