Deep Learning-Based Meat Freshness Detection with Segmentation and OOD-Aware Classification

2026-02-27

AI Summary

This study addresses the challenge of automatically assessing freshness in packaged versus unpackaged meat from RGB images by proposing a robust framework that integrates semantic segmentation with an out-of-distribution (OOD)-aware abstention mechanism. The approach first employs a U-Net to segment meat regions, thereby standardizing the input, and then utilizes multiple backbone networks—including EfficientNet-B0—for classification. An OOD-aware rejection module is incorporated to abstain from prediction on low-confidence or distributionally anomalous samples by returning “no result.” Experimental results demonstrate that the segmentation module achieves an IoU of 75% and a Dice score of 82%, while EfficientNet-B0 attains a classification accuracy of 98.10% on in-distribution test data. The system’s practicality and reliability are further validated through nested cross-validation and successful deployment on mobile devices via TensorFlow Lite.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

In this study, we present a meat freshness classification framework from Red-Green-Blue (RGB) images that supports both packaged and unpackaged meat datasets. The system classifies four in-distribution (ID) meat classes and uses an out-of-distribution (OOD)-aware abstention mechanism that flags low-confidence samples as No Result. The pipeline combines U-Net-based segmentation with deep feature classifiers. Segmentation is used as a preprocessing step to isolate the meat region and reduce background, producing more consistent inputs for classification. The segmentation module achieved an Intersection over Union (IoU) of 75% and a Dice coefficient of 82%, producing standardized inputs for the classification stage. For classification, we benchmark five backbones: Residual Network-50 (ResNet-50), Vision Transformer-Base/16 (ViT-B/16), Swin Transformer-Tiny (Swin-T), EfficientNet-B0, and MobileNetV3-Small. We use nested 5x3 cross-validation (CV) for model selection and hyperparameter tuning. On the held-out ID test set, EfficientNet-B0 achieves the highest accuracy (98.10%), followed by ResNet-50 and MobileNetV3-Small (both 97.63%) and Swin-T (97.51%), while ViT-B/16 is lower (94.42%). We additionally evaluate OOD scoring and thresholding using standard OOD metrics and sensitivity analysis over the abstention threshold. Finally, we report on-device latency using TensorFlow Lite (TFLite) on a smartphone, highlighting practical accuracy-latency trade-offs for future deployment.

Problem

• meat freshness detection
• out-of-distribution detection
• image classification
• food quality assessment
• computer vision

Innovation

• OOD-aware classification
• U-Net segmentation
• meat freshness detection
• cross-validation benchmarking
• on-device inference