Abstract:
The identification of monkey species using automated systems plays a crucial role in
biodiversity research, wildlife conservation, and ecological studies. This study
explores the application of Convolutional Neural Networks (CNNs) for classifying
images of 10 monkey species collected from diverse sources, including the internet,
zoos, and animal houses. A mixed dataset was developed, and six state-of-the-art
CNN architectures—EfficientNet B7, DenseNet 201, VGG19, InceptionV3,
MobileNet, and Xception—were evaluated for their performance. Among these,
DenseNet 201 achieved the highest accuracy of 90.13% with a loss of 1.093,
outperforming MobileNet, which attained 84.94% accuracy with 1.430 loss.
EfficientNet B7 exhibited the lowest performance, with an accuracy of 64.28% and a
loss of 3.76. To demonstrate the practical utility of the best-performing model, a web
interface was developed using Python Flask API, enabling seamless image
classification. This research provides valuable insights into the comparative
performance of CNN architectures for species identification and highlights the
potential of machine learning in advancing wildlife monitoring technologies.Using
the Python Flask API, a web interface was created to show the best-performing
model's usefulness in real-world scenarios by facilitating smooth picture
categorization. This study demonstrates the potential of machine learning to
advance wildlife monitoring technology and offers insightful information on the
relative efficacy of CNN architectures for species identification.
