aura-net : robust segmentation of phase-contrast microscopy images with few annotations

TL;DR

AURA-net is a CNN designed for efficient segmentation of phase-contrast microscopy images, leveraging transfer learning, attention mechanisms, and a specialized loss to perform well with limited annotations.

Contribution

It introduces AURA-net, a novel segmentation network that effectively handles small datasets with limited annotations using transfer learning, attention, and a specialized loss.

Findings

Outperforms state-of-the-art methods on small datasets

Requires fewer annotations for effective training

Achieves high segmentation accuracy with limited data

Abstract

We present AURA-net, a convolutional neural network (CNN) for the segmentation of phase-contrast microscopy images. AURA-net uses transfer learning to accelerate training and Attention mechanisms to help the network focus on relevant image features. In this way, it can be trained efficiently with a very limited amount of annotations. Our network can thus be used to automate the segmentation of datasets that are generally considered too small for deep learning techniques. AURA-net also uses a loss inspired by active contours that is well-adapted to the specificity of phase-contrast images, further improving performance. We show that AURA-net outperforms state-of-the-art alternatives in several small (less than 100images) datasets.