عليرضا صديقي مقدم

Labeled data is a fundamental component in training supervised deep learning models for computer vision tasks. However, the labeling process, especially for ordinal image classification problems where class boundaries are often ambiguous, is prone to error an‎d noise. Such label noise can significantly degrade the performance an‎d reliability of machine learning models. This thesis addresses the problem of detecting an‎d correcting label noise in ordinal image classification tasks. To this end, a novel data-centric method called ORDinal Adaptive Correction (ORDAC) is proposed for adaptive correction of noisy labels. The proposed approach leverages the capabilities of Label Distribution Learning (LDL) to model the inherent ambiguity an‎d uncertainty present in ordinal labels. During training, ORDAC dynamically adjusts the mean an‎d stan‎dard deviation of the label distribution for each sample. Rather than discarding potentially noisy samples, this approach aims to correct them an‎d make optimal use of the entire training dataset. The effectiveness of the proposed method is eva‎luated on benchmark datasets for age estimation (Adience) an‎d disease severity detection (Diabetic Retinopathy) under various asymmetric Gaussian noise scenarios. Results show that ORDAC an‎d its extended versions (ORDAC_C an‎d ORDAC_R) lead to significant improvements in model performance. For instance, on the Adience dataset with 40% noise, ORDAC_R reduced the mean absolute error from 0.86 to 0.62 an‎d increased the recall from 0.37 to 0.49. The method also demonstrated its effectiveness in correcting intrinsic noise present in the original datasets. This research indicates that adaptive label correction using label distributions is an effective strategy to enhance the robustness an‎d accuracy of ordinal classification models in the presence of noisy data.

تشخيص خطاي برچسب , اصلاح برچسب نويزي , دسته‌بندي ترتيبي , نويز برچسب , يادگيري توزيع برچسب , هوش مصنوعي داده‌محور

Label Error Detection , Noisy Label Correction , Ordinal Classification , Label Noise , Label Distribution Learning , Data-centric Artificial Intelligence

Alireza Sedighi Moghaddam

Mohammad Reza Mohammadi