RankED: Addressing Imbalance and Uncertainty in Edge Detection Using Ranking-based Losses

TL;DR

RankED introduces a ranking-based loss method for edge detection that effectively handles class imbalance and label uncertainty, achieving state-of-the-art results on multiple datasets.

Contribution

It proposes a unified ranking-based approach that simultaneously addresses class imbalance and label uncertainty in edge detection tasks.

Findings

Outperforms previous methods on NYUD-v2, BSDS500, and Multi-cue datasets.

Sets new state-of-the-art results in edge detection.

Demonstrates effectiveness of ranking-based losses in handling data imbalance and uncertainty.

Abstract

Detecting edges in images suffers from the problems of (P1) heavy imbalance between positive and negative classes as well as (P2) label uncertainty owing to disagreement between different annotators. Existing solutions address P1 using class-balanced cross-entropy loss and dice loss and P2 by only predicting edges agreed upon by most annotators. In this paper, we propose RankED, a unified ranking-based approach that addresses both the imbalance problem (P1) and the uncertainty problem (P2). RankED tackles these two problems with two components: One component which ranks positive pixels over negative pixels, and the second which promotes high confidence edge pixels to have more label certainty. We show that RankED outperforms previous studies and sets a new state-of-the-art on NYUD-v2, BSDS500 and Multi-cue datasets. Code is available at https://ranked-cvpr24.github.io.