Mini-batch graphs for robust image classification Article Swipe

PDF

Related Concepts

Computer science Artificial intelligence Adversarial system Robustness (evolution) Classifier (UML) Contextual image classification Generative grammar Deep neural networks Graph Pattern recognition (psychology) Artificial neural network Machine learning Image (mathematics) Theoretical computer science Chemistry Biochemistry Gene

Arnab Kumar Mondal , Vineet Jain , Kaleem Siddiqi ·

YOU? · · 2021 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2105.03237 · OA: W3160932900

Current deep learning models for classification tasks in computer vision are trained using mini-batches. In the present article, we take advantage of the relationships between samples in a mini-batch, using graph neural networks to aggregate information from similar images. This helps mitigate the adverse effects of alterations to the input images on classification performance. Diverse experiments on image-based object and scene classification show that this approach not only improves a classifier's performance but also increases its robustness to image perturbations and adversarial attacks. Further, we also show that mini-batch graph neural networks can help to alleviate the problem of mode collapse in Generative Adversarial Networks.