.png){kind=icon}
Return
Decoding class dynamics in learning with noisy labels
.png){kind=icon}
PRE
AI
DOI：10.1016/j.patrec.2024.04.012.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The creation of large-scale datasets annotated by humans inevitably introduces noisy labels, leading to reduced generalization in deep-learning models. Sample selection-based learning with noisy labels is a recent approach that exhibits promising upbeat performance improvements. The selection of clean samples amongst the noisy samples is an important criterion in the learning process of these models. In this work, we delve deeper into the clean-noise split decision and highlight the aspect that effective demarcation of samples would lead to better performance. We identify the Global Noise Conundrum in the existing models, where distribution of samples is treated globally. We propose a per-class-based local distribution of samples and demonstrate the effectiveness of this approach in having a better clean-noise split. We validate our proposal on several benchmarks - both real and synthetic, and show substantial improvements over different stateof-the-art algorithms. We further propose a new metric, classiness to extend our analysis and highlight effectiveness of the proposed method. Source code and instructions to reproduce this paper are available https://github.com/aldakata/CCLM/
Keywords:
Learning with Noisy Labels
Label Noise Modelling
Class Dynamics
Journal
IF:
3.3
Papers:
8.0K
Citations:
1.6W
Organization
Cited Papers
ShareSaveShareSaveLead stimulates Golgi sialyltransferase at times coincident with the embryonic to adult conversion of the neural cell adhesion molecule (N-CAM).png)
Toxicology
ToxicologyIF0
PREAI
ShareSaveShareSaveShareSavePolymorphism for direction of coil in Partula suturalis: Behavioural isolation and positive frequency dependent selectionHeredity
OAAI
HeredityIF0
OAAIShareSave
View more