Multi-Species Object Detection in Drone Imagery for Population Monitoring of Endangered Animals

TL;DR

This paper enhances drone-based animal detection by fine-tuning YOLOv8 models, achieving 95% accuracy in identifying endangered species, enabling real-time monitoring on low-power devices.

Contribution

It introduces a fine-tuning approach for YOLOv8 models with hyperparameter tuning and data augmentation, significantly improving accuracy for drone-based wildlife detection.

Findings

Achieved 95% accuracy on safari animal dataset

Deployed models on Jetson Orin Nano for real-time detection

Improved baseline accuracy from 0.7% to 95%

Abstract

Animal populations worldwide are rapidly declining, and a technology that can accurately count endangered species could be vital for monitoring population changes over several years. This research focused on fine-tuning object detection models for drone images to create accurate counts of animal species. Hundreds of images taken using a drone and large, openly available drone-image datasets were used to fine-tune machine learning models with the baseline YOLOv8 architecture. We trained 30 different models, with the largest having 43.7 million parameters and 365 layers, and used hyperparameter tuning and data augmentation techniques to improve accuracy. While the state-of-the-art YOLOv8 baseline had only 0.7% accuracy on a dataset of safari animals, our models had 95% accuracy on the same dataset. Finally, we deployed the models on the Jetson Orin Nano for demonstration of low-power real-time species detection for easy inference on drones.