# SoilingNet: Soiling Detection on Automotive Surround-View Cameras

## TL;DR

This paper introduces SoilingNet, a CNN-based method for detecting soiling on automotive surround-view cameras, utilizing a new dataset and GAN-based data augmentation to improve autonomous driving safety.

## Contribution

It presents a new dataset with various soiling types, a CNN architecture for detection, and a multi-task framework integrating object detection, along with GAN-based data augmentation.

## Key findings

- High accuracy in soiling detection using CNNs
- Effective multi-task learning with object detection
- GAN-based data augmentation improves model performance

## Abstract

Cameras are an essential part of sensor suite in autonomous driving. Surround-view cameras are directly exposed to external environment and are vulnerable to get soiled. Cameras have a much higher degradation in performance due to soiling compared to other sensors. Thus it is critical to accurately detect soiling on the cameras, particularly for higher levels of autonomous driving. We created a new dataset having multiple types of soiling namely opaque and transparent. It will be released publicly as part of our WoodScape dataset \cite{yogamani2019woodscape} to encourage further research. We demonstrate high accuracy using a Convolutional Neural Network (CNN) based architecture. We also show that it can be combined with the existing object detection task in a multi-task learning framework. Finally, we make use of Generative Adversarial Networks (GANs) to generate more images for data augmentation and show that it works successfully similar to the style transfer.