Delving into Semantic Scale Imbalance
This work addresses model bias for machine learning practitioners by revealing a novel perspective on class imbalance, though it is incremental as it builds on existing long-tailed learning research.
The authors tackled model bias in long-tailed data by defining and quantifying semantic scale imbalance, which explains performance saturation with abundant data, sharp decay with insufficient data, and bias in sample-balanced datasets. Their proposed dynamic semantic-scale-balanced learning achieved superior performance on large-scale long-tailed and non-long-tailed datasets, including natural and medical ones.
Model bias triggered by long-tailed data has been widely studied. However, measure based on the number of samples cannot explicate three phenomena simultaneously: (1) Given enough data, the classification performance gain is marginal with additional samples. (2) Classification performance decays precipitously as the number of training samples decreases when there is insufficient data. (3) Model trained on sample-balanced datasets still has different biases for different classes. In this work, we define and quantify the semantic scale of classes, which is used to measure the feature diversity of classes. It is exciting to find experimentally that there is a marginal effect of semantic scale, which perfectly describes the first two phenomena. Further, the quantitative measurement of semantic scale imbalance is proposed, which can accurately reflect model bias on multiple datasets, even on sample-balanced data, revealing a novel perspective for the study of class imbalance. Due to the prevalence of semantic scale imbalance, we propose semantic-scale-balanced learning, including a general loss improvement scheme and a dynamic re-weighting training framework that overcomes the challenge of calculating semantic scales in real-time during iterations. Comprehensive experiments show that dynamic semantic-scale-balanced learning consistently enables the model to perform superiorly on large-scale long-tailed and non-long-tailed natural and medical datasets, which is a good starting point for mitigating the prevalent but unnoticed model bias.