Multi-Scale Feature Fusion: Learning Better Semantic Segmentation for Road Pothole Detection

TL;DR

This paper introduces a multi-scale feature fusion method for semantic segmentation that improves pothole detection accuracy by integrating spatial context and enhancing feature consistency, achieving state-of-the-art results.

Contribution

The paper proposes a novel multi-scale feature fusion module combined with channel attention and atrous spatial pyramid pooling for improved pothole detection.

Findings

Achieves state-of-the-art performance on Pothole-600 dataset.

Outperforms existing single-modal semantic segmentation networks.

Effective on both RGB and disparity images.

Abstract

This paper presents a novel pothole detection approach based on single-modal semantic segmentation. It first extracts visual features from input images using a convolutional neural network. A channel attention module then reweighs the channel features to enhance the consistency of different feature maps. Subsequently, we employ an atrous spatial pyramid pooling module (comprising of atrous convolutions in series, with progressive rates of dilation) to integrate the spatial context information. This helps better distinguish between potholes and undamaged road areas. Finally, the feature maps in the adjacent layers are fused using our proposed multi-scale feature fusion module. This further reduces the semantic gap between different feature channel layers. Extensive experiments were carried out on the Pothole-600 dataset to demonstrate the effectiveness of our proposed method. The quantitative comparisons suggest that our method achieves the state-of-the-art (SoTA) performance on both RGB images and transformed disparity images, outperforming three SoTA single-modal semantic segmentation networks.