Rethinking Unsupervised Neural Superpixel Segmentation

TL;DR

This paper proposes three improvements to unsupervised CNN-based superpixel segmentation, focusing on image similarity, boundary awareness, and multi-scale architecture, demonstrating enhanced performance on BSDS500.

Contribution

It introduces a novel framework with three key elements—image similarity, boundary consideration, and atrous convolution—to improve unsupervised superpixel segmentation.

Findings

Significant qualitative improvements in superpixel quality.

Quantitative performance gains on BSDS500 dataset.

Enhanced boundary preservation in segmentation results.

Abstract

Recently, the concept of unsupervised learning for superpixel segmentation via CNNs has been studied. Essentially, such methods generate superpixels by convolutional neural network (CNN) employed on a single image, and such CNNs are trained without any labels or further information. Thus, such approach relies on the incorporation of priors, typically by designing an objective function that guides the solution towards a meaningful superpixel segmentation. In this paper we propose three key elements to improve the efficacy of such networks: (i) the similarity of the \emph{soft} superpixelated image compared to the input image, (ii) the enhancement and consideration of object edges and boundaries and (iii) a modified architecture based on atrous convolution, which allow for a wider field of view, functioning as a multi-scale component in our network. By experimenting with the BSDS500 dataset, we find evidence to the significance of our proposal, both qualitatively and quantitatively.