Meta-DRN: Meta-Learning for 1-Shot Image Segmentation

TL;DR

This paper introduces Meta-DRN, a lightweight meta-learning CNN architecture for 1-shot image segmentation that achieves comparable or better performance with significantly fewer parameters.

Contribution

It presents a novel lightweight CNN architecture for 1-shot segmentation, inspired by semantic segmentation models, and demonstrates its effectiveness across multiple meta-learning algorithms.

Findings

70% fewer parameters than benchmarks

Achieves better or comparable mean IoU scores

Effective across four meta-learning algorithms

Abstract

Modern deep learning models have revolutionized the field of computer vision. But, a significant drawback of most of these models is that they require a large number of labelled examples to generalize properly. Recent developments in few-shot learning aim to alleviate this requirement. In this paper, we propose a novel lightweight CNN architecture for 1-shot image segmentation. The proposed model is created by taking inspiration from well-performing architectures for semantic segmentation and adapting it to the 1-shot domain. We train our model using 4 meta-learning algorithms that have worked well for image classification and compare the results. For the chosen dataset, our proposed model has a 70% lower parameter count than the benchmark, while having better or comparable mean IoU scores using all 4 of the meta-learning algorithms.