# Deconstructing the Failure of Ideal Noise Correction: A Three-Pillar Diagnosis Source: [Deconstructing the Failure of Ideal Noise Correction: A Three-Pillar Diagnosis]() Collection index: [llms.txt]() Data updated: 2026-03-17T05:00:04.157Z Authors: Chen Feng, Zhuo Zhi, Zhao Huang, Jiawei Ge, Ling Xiao, Nicu Sebe, Georgios Tzimiropoulos, Ioannis Patras Published: 2026-03-13T13:53:04.000Z ## Abstract Statistically consistent methods based on the noise transition matrix ($T$) offer a theoretically grounded solution to Learning with Noisy Labels (LNL), with guarantees of convergence to the optimal clean-data classifier. In practice, however, these methods are often outperformed by empirical approaches such as sample selection, and this gap is usually attributed to the difficulty of accurately estimating $T$. The common assumption is that, given a perfect $T$, noise-correction methods would recover their theoretical advantage. In this work, we put this longstanding hypothesis to a decisive test. We conduct experiments under idealized conditions, providing correction methods with a perfect, oracle transition matrix. Even under these ideal conditions, we observe that these methods still suffer from performance collapse during training. This compellingly demonstrates that the failure is not fundamentally a $T$-estimation problem, but stems from a more deeply rooted flaw. To explain this behaviour, we provide a unified analysis that links three levels: macroscopic convergence states, microscopic optimisation dynamics, and information-theoretic limits on what can be learned from noisy labels. Together, these results give a formal account of why ideal noise correction fails and offer concrete guidance for designing more reliable methods for learning with noisy labels. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Classification on CIFAR-10 (test) (Symmetric/Asymmetric Accuracy Benchmarking)](): Image Classification; CIFAR-10 (test); Accuracy (Symmetric 20%): 95.6. - [Image Classification on CIFAR-100 (test) (Symmetric Noise Accuracy 20/50/80/90%)](): Image Classification; CIFAR-100 (test); Accuracy (Symmetric 20%): 73.1. - [Image Classification on Clothing1M (test)](): Image Classification; Clothing1M (test); Accuracy: 72.24. ## Original sources and implementation [Original paper]() [PDF]()