Mitigating the Bias of Centered Objects in Common Datasets

2021/12/16 by Gergely Szabó, Szabo, Gergely, András Horváth +1 · 2 citations
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Machine Learning (cs.LG) #Neural Networks and Applications

paper · pdf · doi:10.48550/arxiv.2112.09195

openalex publication_date 2021/12/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

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.

Cited by

• Spanning Tree Autoregressive Visual Generation
• Classifying polish in use-wear analysis with convolutional neural networks

Related

• Mind the Pad -- CNNs can Develop Blind Spots
• Assessing Shape Bias Property of Convolutional Neural Networks
• Corner Cases: How Size and Position of Objects Challenge ImageNet-Trained Models
• An Empirical Method to Quantify the Peripheral Performance Degradation in Deep Networks
• An Empirical Evaluation of Current Convolutional Architectures' Ability to Manage Nuisance Location and Scale Variability