Towards Measuring Bias in Image Classification
This work addresses bias detection for improving trust in CNNs in industrial applications, but it is incremental as it builds on existing attribution methods.
The paper tackles the problem of detecting data bias in image classification by proposing a systematic approach using attribution maps on artificially biased datasets, showing that some techniques and metrics can effectively highlight bias.
Convolutional Neural Networks (CNN) have become de fact state-of-the-art for the main computer vision tasks. However, due to the complex underlying structure their decisions are hard to understand which limits their use in some context of the industrial world. A common and hard to detect challenge in machine learning (ML) tasks is data bias. In this work, we present a systematic approach to uncover data bias by means of attribution maps. For this purpose, first an artificial dataset with a known bias is created and used to train intentionally biased CNNs. The networks' decisions are then inspected using attribution maps. Finally, meaningful metrics are used to measure the attribution maps' representativeness with respect to the known bias. The proposed study shows that some attribution map techniques highlight the presence of bias in the data better than others and metrics can support the identification of bias.