miscellaneous-8

Label Error Detection

3260 papers • 126 benchmarks • 313 datasets

Identify labeling errors in data

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in label-error-detection-8

Trend

Dataset

Best Model

Actions

TREC-6

Libraries

i

Use these libraries to find label-error-detection-8 models and implementations

Datasets

No datasets available.

Subtasks

No subtasks available.

Most implemented papers

Identifying Incorrect Annotations in Multi-Label Classification Data

Curtis G. Northcutt, Jonas W. Mueller, Aditya Thyagarajan, El'ias Snorrason•Thu Nov 24 2022

An extension of the Confident Learning framework is proposed to this setting, as well as a label quality score that ranks examples with label errors much higher than those which are correctly labeled.

15

Content

0

Paper Graph

WENETSPEECH: A 10000+ Hours Multi-Domain Mandarin Corpus for Speech Recognition

Hui Bu, Xin Xu, Lei Xie, Binbin Zhang, Hang Lv, Pengcheng Guo, Qijie Shao, Chao Yang, Xiaoyu Chen, Chenchen Zeng, Di Wu, Zhendong Peng•Wed Oct 06 2021

WenetSpeech is the current largest open-source Mandarin speech corpus with transcriptions, which benefits research on production-level speech recognition, and a novel end-to-end label error detection approach is proposed.

296 0

Paper Graph

CTRL: Clustering Training Losses for Label Error Detection

N. Jha, C. Yue•Tue Aug 16 2022

This work proposes a novel framework, called CTRL11CTRL (Clustering TRaining Losses for label error detection), to detect label errors in multiclass datasets, and demonstrates state-of-the-art error detection accuracy on both image and tabular datasets under labeling noise.

22 0

Paper Graph

Automated Detection of Label Errors in Semantic Segmentation Datasets via Deep Learning and Uncertainty Quantification

M. Rottmann, Marco Reese•Tue Jul 12 2022

This work presents for the first time a method for detecting label errors in image datasets with semantic segmentation, i.e., pixel-wise class labels by lifting the consideration of uncertainty to the level of predicted components and enabling the usage of DNNs together with component-level uncertainty quantification for the detection of label errors.

34 0

Paper Graph

The Re-Label Method For Data-Centric Machine Learning

Tong Guo•Wed Feb 08 2023

The idea for a broad set of deep learning tasks, includes classification, sequence tagging, object detection, sequence generation, click-through rate prediction, is illustrated.

2 0

Paper Graph

AQuA: A Benchmarking Tool for Label Quality Assessment

Arjun Choudhry, Mononito Goswami, Vedant Sanil, Arvind Srinivasan, Chalisa Udompanyawit, A. Dubrawski•Wed Jun 14 2023

A benchmarking environment AQuA is proposed to rigorously evaluate methods that enable machine learning in the presence of label noise and a design space is introduced to delineate concrete design choices of label error detection models.

18 0

Paper Graph

Adding a benchmark result helps the community track progress.