Multi-Task Classification of Sewer Pipe Defects and Properties using a Cross-Task Graph Neural Network Decoder

TL;DR

This paper introduces a novel multi-task graph neural network architecture, CT-GNN, that simultaneously classifies sewer pipe defects and properties, improving accuracy and efficiency over existing methods.

Contribution

The paper proposes the Cross-Task Graph Neural Network (CT-GNN), a new multi-task classification architecture that refines predictions using cross-task information and can be integrated into any backbone.

Findings

Achieves state-of-the-art performance on Sewer-ML dataset tasks.

Improves defect and water level classification accuracy by 5.3 and 8.0 percentage points.

Uses 50 times fewer parameters than previous models.

Abstract

The sewerage infrastructure is one of the most important and expensive infrastructures in modern society. In order to efficiently manage the sewerage infrastructure, automated sewer inspection has to be utilized. However, while sewer defect classification has been investigated for decades, little attention has been given to classifying sewer pipe properties such as water level, pipe material, and pipe shape, which are needed to evaluate the level of sewer pipe deterioration. In this work we classify sewer pipe defects and properties concurrently and present a novel decoder-focused multi-task classification architecture Cross-Task Graph Neural Network (CT-GNN), which refines the disjointed per-task predictions using cross-task information. The CT-GNN architecture extends the traditional disjointed task-heads decoder, by utilizing a cross-task graph and unique class node embeddings. The cross-task graph can either be determined a priori based on the conditional probability between the task classes or determined dynamically using self-attention. CT-GNN can be added to any backbone and trained end-to-end at a small increase in the parameter count. We achieve state-of-the-art performance on all four classification tasks in the Sewer-ML dataset, improving defect classification and water level classification by 5.3 and 8.0 percentage points, respectively. We also outperform the single task methods as well as other multi-task classification approaches while introducing 50 times fewer parameters than previous model-focused approaches. The code and models are available at the project page http://vap.aau.dk/ctgnn