Illumination-Based Data Augmentation for Robust Background Subtraction

TL;DR

This paper introduces a novel data augmentation technique that simulates illumination changes like flashes and shadows to improve the robustness of deep learning models for background subtraction in videos.

Contribution

It proposes a new illumination-based data augmentation method that enhances model generalization to sudden lighting changes in background subtraction tasks.

Findings

Augmentation improves model performance under illumination changes.

Synthetic data helps train illumination-invariant deep learning models.

Method effectively simulates real-world lighting variations.

Abstract

A core challenge in background subtraction (BGS) is handling videos with sudden illumination changes in consecutive frames. In this paper, we tackle the problem from a data point-of-view using data augmentation. Our method performs data augmentation that not only creates endless data on the fly, but also features semantic transformations of illumination which enhance the generalisation of the model. It successfully simulates flashes and shadows by applying the Euclidean distance transform over a binary mask that is randomly generated. Such data allows us to effectively train an illumination-invariant deep learning model for BGS. Experimental results demonstrate the contribution of the synthetics in the ability of the models to perform BGS even when significant illumination changes take place. The source code of the project is made publicly available at https://github.com/dksakkos/illumination_augmentation.