Objective 1: Develop benchmark ACP dataset
Create a specialized image dataset of Asian Citrus Psyllid (ACP) to train and validate AI models, enhancing accuracy in small-target detection
Objective 2: Build AI model for pest identification
Design and optimize deep learning architectures (e.g., YOLO, Faster R-CNN) to detect ACP on sticky traps with minimal human intervention
Objective 3: Deploy AI-integrated early warning system
Combine the trained AI model with automated traps and wireless communication for real-time field monitoring, validated in lab settings
Objective 4: Uncover environmental drivers of ACP outbreaks
Apply causal machine learning to analyze how weather (temperature, rainfall) and satellite-derived variables (NDVI, soil moisture) influence pest populations