Search papers, labs, and topics across Lattice.
FryNet, a dual-stream RGB-thermal framework, was developed to non-destructively assess frying oil oxidation by jointly performing oil-region segmentation, serviceability classification, and chemical oxidation index regression. The key innovation is a Dual-Encoder DANN that adversarially regularizes both streams against video identity, preventing the model from exploiting camera-fingerprint shortcuts and forcing it to learn chemically grounded representations. FryNet achieves state-of-the-art performance on a dataset of 7,226 paired frames, demonstrating its ability to accurately assess oil degradation without destructive methods.
By adversarially removing camera-specific fingerprints, FryNet forces models to learn genuine chemical representations from thermal images, enabling robust and generalizable frying oil oxidation assessment.
Monitoring frying oil degradation is critical for food safety, yet current practice relies on destructive wet-chemistry assays that provide no spatial information and are unsuitable for real-time use. We identify a fundamental obstacle in thermal-image-based inspection, the camera-fingerprint shortcut, whereby models memorize sensor-specific noise and thermal bias instead of learning oxidation chemistry, collapsing under video-disjoint evaluation. We propose FryNet, a dual-stream RGB-thermal framework that jointly performs oil-region segmentation, serviceability classification, and regression of four chemical oxidation indices (PV, p-AV, Totox, temperature) in a single forward pass. A ThermalMiT-B2 backbone with channel and spatial attention extracts thermal features, while an RGB-MAE Encoder learns chemically grounded representations via masked autoencoding and chemical alignment. Dual-Encoder DANN adversarially regularizes both streams against video identity via Gradient Reversal Layers, and FiLM fusion bridges thermal structure with RGB chemical context. On 7,226 paired frames across 28 frying videos, FryNet achieves 98.97% mIoU, 100% classification accuracy, and 2.32 mean regression MAE, outperforming all seven baselines.
