Projects
Vision Language Foundation Models for Plant Simulation
A novel vision language framework utilizing in-context learning to generate 3D plant simulation JSON configurations directly from drone-based remote sensing images.
WeedTrackr: Precision Weed Management
Our 1st Place winning entry for the 2025 AIFS × Sony AI AgTech Challenge. A low-cost Edge AI weed monitoring system executing custom deep learning models compiled for the Raspberry Pi AI Camera.
VisTA-SR: Thermal Camera Super-Resolution
A dual-stage multimodal deep learning framework designed to enhance both the spatial resolution and temperature accuracy of low-cost consumer thermal cameras using RGB-guided fusion for precision agriculture.
Multimodal Optical Biosensing via 3D-CNN
A non-destructive phenotyping approach using a 3D-CNN model to merge RGB, Depth, and Chlorophyll Fluorescence data. This framework accurately classifies water deficit stress in basil with 96.9% accuracy.
Robo-ag: Targeted Spray Application
Winner of the Elegance in Design Prize at the 2023 Farm Robotics Challenge. An autonomous targeted pesticide applicator utilizing the farm-ng Amiga platform to reduce chemical waste in vineyards.