Neural Relation Graph: A Unified Framework for Identifying Label Noise and Outlier Data
This addresses the challenge of diagnosing and cleaning data for robust machine learning systems across domains like image, speech, and language, though it appears incremental as it builds on existing relational graph concepts.
The paper tackles the problem of identifying label noise and outlier data in large-scale datasets by proposing a unified framework based on relational graph structures in feature-embedded space, achieving state-of-the-art detection performance on tasks like ImageNet, ESC-50, and SST2.
Diagnosing and cleaning data is a crucial step for building robust machine learning systems. However, identifying problems within large-scale datasets with real-world distributions is challenging due to the presence of complex issues such as label errors, under-representation, and outliers. In this paper, we propose a unified approach for identifying the problematic data by utilizing a largely ignored source of information: a relational structure of data in the feature-embedded space. To this end, we present scalable and effective algorithms for detecting label errors and outlier data based on the relational graph structure of data. We further introduce a visualization tool that provides contextual information of a data point in the feature-embedded space, serving as an effective tool for interactively diagnosing data. We evaluate the label error and outlier/out-of-distribution (OOD) detection performances of our approach on the large-scale image, speech, and language domain tasks, including ImageNet, ESC-50, and SST2. Our approach achieves state-of-the-art detection performance on all tasks considered and demonstrates its effectiveness in debugging large-scale real-world datasets across various domains. We release codes at https://github.com/snu-mllab/Neural-Relation-Graph.