Mitigating the Bias of Centered Objects in Common Datasets
This addresses a subtle but impactful bias in computer vision datasets that affects model robustness, though the solution is incremental.
The paper identified a bias in common datasets where objects are over-represented at image centers, causing significant accuracy drops as objects approach boundaries, and showed that data augmentation can mitigate this effect.
Convolutional networks are considered shift invariant, but it was demonstrated that their response may vary according to the exact location of the objects. In this paper we will demonstrate that most commonly investigated datasets have a bias, where objects are over-represented at the center of the image during training. This bias and the boundary condition of these networks can have a significant effect on the performance of these architectures and their accuracy drops significantly as an object approaches the boundary. We will also demonstrate how this effect can be mitigated with data augmentation techniques.