Transfer Learning for Mosquito Classification Using VGG16
Ayesha Anam Irshad Siddiqui, Charansing Kayte · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2023
A challenge in computer vision known mosquito classification hasn't gained much traction.Automatic mosquito species credentials using real-time images is a crucial feature.Mosquitoes are a serious matter of concern since they can spread diseases including dengue fever, zika, and malaria.It's important to control mosquito populations in order to effectively control mosquitoes.The World Health Organization reported that over a million people worldwide experience malaria and dengue fever each year.In this investigation, we analyze a deep learning vgg-16 network architecture for mosquito specifically chosen.On our mosquito dataset, which included six (6) species of mosquito.The pretrained vgg-16 network architecture with transfer learning technique was studied and proved to identify distinct mosquito species, with an average accuracy rate of 97.1751 percent Loss 0.094359393954277.The results of VGG 16 and CNN are compared.The results show that CNN with multi class classifier is achieving 85.75 percent accuracy and VGG 16 with 97.1751 accuracy.It shows that the VGG 16 model is pretty good in results as compare to CNN.