DRBANET: A Lightweight Dual-Resolution Network for Semantic Segmentation with Boundary Auxiliary

TL;DR

DRBANet is a lightweight dual-resolution network that enhances semantic segmentation by incorporating boundary information, achieving a good balance of accuracy and efficiency on urban scene datasets.

Contribution

The paper introduces DRBANet, a novel dual-resolution architecture with boundary supervision, improving segmentation quality in lightweight models.

Findings

Achieves state-of-the-art accuracy on Cityscapes and CamVid datasets.

Maintains high efficiency suitable for real-time applications.

Effectively captures boundary details to refine segmentation results.

Abstract

Due to the powerful ability to encode image details and semantics, many lightweight dual-resolution networks have been proposed in recent years. However, most of them ignore the benefit of boundary information. This paper introduces a lightweight dual-resolution network, called DRBANet, aiming to refine semantic segmentation results with the aid of boundary information. DRBANet adopts dual parallel architecture, including: high resolution branch (HRB) and low resolution branch (LRB). Specifically, HRB mainly consists of a set of Efficient Inverted Bottleneck Modules (EIBMs), which learn feature representations with larger receptive fields. LRB is composed of a series of EIBMs and an Extremely Lightweight Pyramid Pooling Module (ELPPM), where ELPPM is utilized to capture multi-scale context through hierarchical residual connections. Finally, a boundary supervision head is designed to capture object boundaries in HRB. Extensive experiments on Cityscapes and CamVid datasets demonstrate that our method achieves promising trade-off between segmentation accuracy and running efficiency.