Carly Chery
Deep Learning Researcher · Computer Vision for Plants · Imaging & Sensing Engineering
Final-year B.Sc. student, Agricultural Sciences
EARTH University · Guácimo, Costa Rica
Graduation thesis supervisor: Dr. Víctor Hugo Morales Peña
I am a final-year agricultural sciences student at EARTH University in Costa Rica, working in deep learning and computer vision for plants. I turn plant images and sensor data into measurements a grower can act on, and I build the imaging and sensing hardware those models depend on.
Research interests
- Computer vision for plants
- Image-based plant phenotyping
- Plant disease detection & severity grading
- Imaging & sensing instrumentation
- Controlled-environment agriculture
- Edge / on-device deep learning
Research areas
Research statement and projects →
Plant-disease vision
Deep-learning segmentation models that measure coffee leaf rust and ojo de gallo severity and grade it G0 to G3; the compact ones run offline in a phone app.
Read more →
Imaging rig for deep learning
A 3D-printed imaging rig I designed and built to help people collect consistent images for training deep-learning models.
Read more →
Plant sensing & controlled environments
Sensor data-logging for a tilapia and greenhouse-tomato aquaponic system at Auburn University, and an Arduino controller that regulates hydroponic pH.
Read more →Selected publications
-
Leakage-Aware Temporal Validation of Machine-Learning Models for Yield Prediction in Coffea arabica var. Esperanza L4A5 Under Lowland Agroforestry Systems
Preprint on Preprints.org · submitted to Agriculture (MDPI) · DOI ↗ -
Comparative evaluation of three semantic segmentation architectures for severity assessment of Hemileia vastatrix and Mycena citricolor on Coffea arabica var. Esperanza L4A5
EARTH University, Costa Rica · Code ↗ · Zenodo DOI ↗
News
-
Oct 2026
Co-authored preprint on leakage-aware temporal validation of machine-learning models for coffee yield prediction, now on Preprints.org and submitted to Agriculture (MDPI).
-
Oct 2026
Thesis: under one matched training protocol, a compact UNet++ came within about 0.01 leaf IoU of a SegFormer-B5 transformer twelve times its size, and ojo de gallo severity was predicted within 0.3 %.
-
Sep – Dec 2025
Research internship at Auburn University's E.W. Shell Fisheries Center, building the sensor data-logging workflow for an integrated aquaponics system.
-
Feb 2024
Delegate at the HPAIR Harvard Conference, Harvard University, Cambridge, MA.
-
2024
Our Arduino pH controller for hydroponics placed 6th of 16 at the Cenfotec inter-university competition, Costa Rica.
-
2023
Began the B.Sc. in Agricultural Sciences at EARTH University, Costa Rica.