FruitProM-V2: Robust Probabilistic Maturity Estimation and Detection of Fruits and Vegetables

2026-04-28

AI Summary

This work addresses the limitations of conventional multi-class classification approaches for fruit maturity detection, which struggle with ambiguous boundaries between adjacent stages and inconsistent labeling. The study proposes a novel formulation that models maturity as a latent continuous variable and introduces a probabilistic detection head grounded in distributional representation. By leveraging the cumulative distribution function (CDF) to map this continuous representation to categorical probabilities, the method more effectively captures the inherent uncertainty in the ripening process. Integrated into a deep object detection framework, the approach achieves performance on par with baseline methods under clean labels and demonstrates significantly superior robustness in the presence of label noise, thereby validating the efficacy of the proposed probabilistic modeling strategy in enhancing both representational capacity and resilience to annotation inconsistencies.

Abstract

Accurate fruit maturity identification is essential for determining harvest timing, as incorrect assessment directly affects yield and post-harvest quality. Although ripening is a continuous biological process, vision-based maturity estimation is typically formulated as a multi-class classification task, which imposes sharp boundaries between visually similar stages. To examine this limitation, we perform an annotation reliability study with two independent annotators on a held-out tomato dataset and observe disagreement concentrated near adjacent maturity stages. Motivated by this observation, we model maturity as a latent continuous variable and predict it probabilistically using a distributional detection head, converting the distribution into class probabilities through the cumulative distribution function (CDF). The proposed formulation maintains comparable performance to a standard detector under clean labels while better representing uncertainty. Furthermore, when controlled label noise is introduced during training, the probabilistic model demonstrates improved robustness relative to the baseline, indicating that explicitly modeling maturity uncertainty leads to more reliable visual maturity estimation.

Problem

• fruit maturity estimation
• probabilistic modeling
• label uncertainty
• continuous maturity
• robustness

Innovation

• probabilistic maturity estimation
• continuous maturity modeling
• distributional detection head
• label noise robustness
• cumulative distribution function (CDF)