Making Every Label Count: Handling Semantic Imprecision by Integrating Domain Knowledge

TL;DR

This paper introduces CHILLAX, a hierarchical classification method that effectively leverages imprecise labels in noisy datasets, outperforming existing approaches in image classification tasks.

Contribution

The paper proposes CHILLAX, a novel hierarchical classification approach that handles label imprecision, improving learning from weakly labeled data.

Findings

Outperforms strong baselines by up to 16.4 percentage points.

Surpasses current state-of-the-art by up to 3.9 percentage points.

Effective on noisy variants of NABirds and ILSVRC2012 datasets.

Abstract

Noisy data, crawled from the web or supplied by volunteers such as Mechanical Turkers or citizen scientists, is considered an alternative to professionally labeled data. There has been research focused on mitigating the effects of label noise. It is typically modeled as inaccuracy, where the correct label is replaced by an incorrect label from the same set. We consider an additional dimension of label noise: imprecision. For example, a non-breeding snow bunting is labeled as a bird. This label is correct, but not as precise as the task requires. Standard softmax classifiers cannot learn from such a weak label because they consider all classes mutually exclusive, which non-breeding snow bunting and bird are not. We propose CHILLAX (Class Hierarchies for Imprecise Label Learning and Annotation eXtrapolation), a method based on hierarchical classification, to fully utilize labels of any precision. Experiments on noisy variants of NABirds and ILSVRC2012 show that our method outperforms strong baselines by as much as 16.4 percentage points, and the current state of the art by up to 3.9 percentage points.