“…MTL aims to improve the performance of multiple related tasks by simultaneous training. Related work has proven MTL's effectiveness in various image classification tasks, e.g., face recognition ( Lu et al, 2023 ), medical imaging diagnosis ( Zhao et al, 2023 ; Tang et al, 2023 ; Sun et al, 2023 ; Zhou et al, 2021 ) , and food safety ( Amrani et al, 2024 ). However, research on the application of multi-task learning in fruit freshness detection is still underdeveloped and requires further exploration.…”