Enhancing Jujube Forest Growth Estimation and Disease Detection Using a Novel Diffusion-Transformer Architecture

Area of Science:

• Forestry science
• Computer science
• Artificial intelligence

Background:

• Current forestry monitoring methods struggle with large-scale, complex forest areas due to data processing and feature extraction limitations.
• Accurate assessment of tree health and growth is crucial for sustainable forest management.

Purpose of the Study:

• To propose and evaluate an advanced deep learning model for jujube forest growth estimation and disease detection.
• To address the limitations of existing forestry monitoring techniques.

Main Methods:

• Developed a deep learning model integrating Diffusion-Transformer structure and parallel attention mechanism.
• Conducted benchmark tests and ablation experiments to evaluate performance using precision, recall, accuracy, and F1-score.
• Compared the proposed model against Support Vector Machines, Random Forests, AlexNet, and ResNet.

Main Results:

• The proposed model achieved 95% precision, 92% recall, 93% accuracy, and 94% F1-score for disease detection.
• Demonstrated superior performance in growth estimation tasks compared to traditional and common deep learning models.
• Ablation studies confirmed the effectiveness of parallel attention and parallel loss functions.

Conclusions:

• The proposed model offers a significant advancement for forestry disease monitoring and health assessment.
• Provides a new technical approach and theoretical foundation for intelligent forestry management.
• Highlights the potential of advanced deep learning architectures in ecological applications.