Uncertainty Estimation and Calibration of a Few-Shot Transfer Learning Model for Lettuce Phenotyping

Rusith Chamara Hathurusinghe Dewage, Rusith Chamara Hathurusinghe Dewage, Habib Ullah, Muhammad Salman. Siddiqui, Rakibul Islam, Fadi-Al Machot

2025

Abstract

Computer vision-assisted automatic plant phenotyping in controlled environment agriculture (CEA) remains a significant challenge due to the scarcity of labeled data from growing conditions. In this work, we investigate few-shot transfer learning for estimating the maximum width of lettuce from cropped and segmented images exhibiting non-uniform spatial distribution. The dataset presents additional complexity as images are captured using a wide-angle, off-center camera. We systematically investigate backbone architectures (ResNet, EfficientNet, MobileNet, DenseNet, and Vision Transformer) and perform various data augmentation strategies and regression head designs to identify optimal configurations under few-shot conditions. To enhance predictive reliability, we employ post-hoc uncertainty estimation using Monte Carlo (MC) dropout and conformal prediction, and further evaluate model calibration to analyze alignment between predicted uncertainties and empirical errors. Our best model, based on Vision Transformer Huge with 14x14 patch size (ViT-H/14), achieved a root mean square error (RMSE) of 14.34 mm on the test set. For uncertainty estimation, MC dropout achieved a miscalibration area of 0.19, an average prediction interval width of 27.89 mm, and an empirical coverage of 73\% at the nominal 90\% confidence level. Our results highlight the importance of backbone selection, augmentation, and head architecture on model generalization and reliability. This study offers practical guidelines for developing robust, uncertainty-aware few-shot models for plant phenotyping, enabling more trustworthy deployment in CEA applications.

Paper Citation

in Harvard Style

Dewage R., Ullah H., Siddiqui M., Islam R. and Machot F. (2025). Uncertainty Estimation and Calibration of a Few-Shot Transfer Learning Model for Lettuce Phenotyping. In Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - Volume 1: KDIR; ISBN , SciTePress, pages 388-397. DOI: 10.5220/0013743600004000

in Bibtex Style

@conference{kdir25,
author={Rusith Dewage and Habib Ullah and Muhammad Siddiqui and Rakibul Islam and Fadi-Al Machot},
title={Uncertainty Estimation and Calibration of a Few-Shot Transfer Learning Model for Lettuce Phenotyping},
booktitle={Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - Volume 1: KDIR},
year={2025},
pages={388-397},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013743600004000},
isbn={},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - Volume 1: KDIR
TI - Uncertainty Estimation and Calibration of a Few-Shot Transfer Learning Model for Lettuce Phenotyping
SN -
AU - Dewage R.
AU - Ullah H.
AU - Siddiqui M.
AU - Islam R.
AU - Machot F.
PY - 2025
SP - 388
EP - 397
DO - 10.5220/0013743600004000
PB - SciTePress