Background-based augmentation improves electron microscopy virus detection
Bg-Mix: On Using Image Backgrounds for Data Augmentation in Electron Microscopy Images
Computer Vision and Pattern Recognition
Summary
Electron microscopy images are used to study viruses, but there are usually not enough labeled images to train strong AI models. The authors came up with a way to mix backgrounds and virus images to create more varied training pictures. This helps AI systems get better at finding viruses in new images. Their method improved detection accuracy on two different image sets and two AI models.
What this means in practice
- •For medical imaging teams: Improve virus detection accuracy in electron microscopy images by augmenting training data with varied backgrounds.
- •For microscopy software developers: Enhance object detection tools for pathogen analysis by integrating background-based image augmentation techniques.
Authors
Ankur Srivastava, Joana Sarah Grah, Katharina Ladewig, Nils Körber
Abstract
Electron microscopy (EM) serves as a cornerstone in virology enabling the visualization of viral particles and is widely used in research, diagnostics and morphological characterization of pathogens. Computer-assisted image analysis can help facilitate and speed up this labor-intensive method, however, typical EM virus datasets are small and sparsely labeled, resulting in a lack of readily available deep learning models due to training limitations. To overcome data shortages and leverage content-rich, diverse sample backgrounds, we employed background-based data augmentation. To this end, we propose Bg-Mix, a form of data augmentation method, which decomposes training images into foreground objects and backgrounds and subsequently rearranges both into novel augmented images. This not only enhances the performance of object detection models but also increases the diversity of the training data. Using Bg-Mix provides an increase between +0.8 and +3.3 box mean average precision (mAP) points over the baseline when tested across two datasets using two architectures, thereby demonstrating its effectiveness.