Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Centrality and consistency: two-stage clean samples identification for learning with instance-dependent noisy labels

Zhao, Ganlong, Li, Guanbin, Qin, Yipeng ORCID: https://orcid.org/0000-0002-1551-9126, Liu, Feng and Yu, Yizhou 2022. Centrality and consistency: two-stage clean samples identification for learning with instance-dependent noisy labels. Presented at: European Conference on Computer Vision (ECCV 2022), Tel Aviv, Israel, 23-27 October 2022. Published in: Avidan, Shai, Brostow, Gabriel, Cisse, Moustapha, Farinella, Giovanni Maria and Hassner, Tal eds. Proceedings of the Computer Vision – ECCV 2022. IEEE, pp. 21-37. 10.1007/978-3-031-19806-9_2

[thumbnail of Centrality and Consistency_ECCV2022.pdf]
Preview
PDF - Accepted Post-Print Version
Download (1MB) | Preview

Abstract

Deep models trained with noisy labels are prone to over-fitting and struggle in generalization. Most existing solutions are based on an ideal assumption that the label noise is class-conditional, i.e. instances of the same class share the same noise model, and are independent of features. While in practice, the real-world noise patterns are usually more fine-grained as instance-dependent ones, which poses a big challenge, especially in the presence of inter-class imbalance. In this paper, we propose a two-stage clean samples identification method to address the aforementioned challenge. First, we employ a class-level feature clustering procedure for the early identification of clean samples that are near the class-wise prediction centers. Notably, we address the class imbalance problem by aggregating rare classes according to their prediction entropy. Second, for the remaining clean samples that are close to the ground truth class boundary (usually mixed with the samples with instance-dependent noises), we propose a novel consistency-based classification method that identifies them using the consistency of two classifier heads: the higher the consistency, the larger the probability that a sample is clean. Extensive experiments on several challenging benchmarks demonstrate the superior performance of our method against the state-of-the-art. Code is available at https://github.com/uitrbn/TSCSI_IDN.

Item Type: Conference or Workshop Item (Paper)
Date Type: Published Online
Status: Published
Schools: Schools > Computer Science & Informatics
Publisher: IEEE
ISBN: 978-3-031-19805-2
Date of First Compliant Deposit: 25 July 2022
Last Modified: 13 May 2025 15:00
URI: https://orca.cardiff.ac.uk/id/eprint/151373

Actions (repository staff only)

Edit Item Edit Item

Downloads

Downloads per month over past year

View more statistics