Efficient Classification of Very Large Images with Tiny Objects

TL;DR

This paper introduces a memory-efficient CNN called Zoom-In network that effectively classifies large images with tiny objects by using hierarchical attention sampling, outperforming existing methods across multiple datasets.

Contribution

The paper proposes a novel end-to-end CNN architecture that handles ultra-large images with small informative regions using hierarchical attention, requiring only a single GPU.

Findings

Higher accuracy than existing methods on multiple datasets

Requires less memory resources

Effective on gigapixel pathology images

Abstract

An increasing number of applications in computer vision, specially, in medical imaging and remote sensing, become challenging when the goal is to classify very large images with tiny informative objects. Specifically, these classification tasks face two key challenges: $i$) the size of the input image is usually in the order of mega- or giga-pixels, however, existing deep architectures do not easily operate on such big images due to memory constraints, consequently, we seek a memory-efficient method to process these images; and $ii$) only a very small fraction of the input images are informative of the label of interest, resulting in low region of interest (ROI) to image ratio. However, most of the current convolutional neural networks (CNNs) are designed for image classification datasets that have relatively large ROIs and small image sizes (sub-megapixel). Existing approaches have addressed these two challenges in isolation. We present an end-to-end CNN model termed Zoom-In network that leverages hierarchical attention sampling for classification of large images with tiny objects using a single GPU. We evaluate our method on four large-image histopathology, road-scene and satellite imaging datasets, and one gigapixel pathology dataset. Experimental results show that our model achieves higher accuracy than existing methods while requiring less memory resources.