Transfer Learning for Wildlife Classification: Evaluating YOLOv8 against DenseNet, ResNet, and VGGNet on a Custom Dataset

Subek Sharma,Sisir Dhakal,Mansi Bhavsar

DOI: https://doi.org/10.36548/jaicn.2024.4.003

2024-11-12

Abstract:This study evaluates the performance of various deep learning models, specifically DenseNet, ResNet, VGGNet, and YOLOv8, for wildlife species classification on a custom dataset. The dataset comprises 575 images of 23 endangered species sourced from reputable online repositories. The study utilizes transfer learning to fine-tune pre-trained models on the dataset, focusing on reducing training time and enhancing classification accuracy. The results demonstrate that YOLOv8 outperforms other models, achieving a training accuracy of 97.39% and a validation F1-score of 96.50%. These findings suggest that YOLOv8, with its advanced architecture and efficient feature extraction capabilities, holds great promise for automating wildlife monitoring and conservation efforts.

Computer Vision and Pattern Recognition,Artificial Intelligence

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the accuracy and efficiency of wildlife species classification, especially for endangered species. Current wildlife monitoring methods usually rely on images captured by camera traps, and these images need to be analyzed manually. This method is time - consuming, cumbersome and error - prone. By introducing deep - learning techniques, especially Convolutional Neural Networks (CNNs), researchers hope to improve the automation level of species identification, thereby reducing manual labor and improving monitoring efficiency and accuracy.

Specifically, the paper evaluates the performance of several different deep - learning models on a custom - made dataset. These models include DenseNet, ResNet, VGGNet and YOLOv8. The main objective of the study is to determine which model can provide the best classification accuracy and efficiency, especially when processing images of endangered species. In addition, the study also explores the challenges and limitations that these models may encounter in practical applications, aiming to provide guidance for researchers and practitioners when choosing models suitable for endangered species protection tasks.

Through this study, the author hopes to fill the gaps existing in current wildlife monitoring methods, that is, by systematically evaluating different deep - learning models and providing crucial guidance to promote more efficient and accurate wildlife monitoring and protection work.