Mark My Words: Dangers of Watermarked Images in ImageNet

TL;DR

This paper investigates how watermarks in ImageNet images influence pre-trained neural networks, revealing widespread effects across multiple classes and proposing a mitigation method to reduce watermark-related artifacts in feature representations.

Contribution

It demonstrates that watermarks affect various ImageNet classes and introduces a simple technique to mitigate watermark artifacts in fine-tuned models.

Findings

Watermarks influence multiple ImageNet classes beyond the 'carton' class.

Pre-trained networks encode watermark patterns in their features.

Ignoring watermark-affected features improves model robustness.

Abstract

The utilization of pre-trained networks, especially those trained on ImageNet, has become a common practice in Computer Vision. However, prior research has indicated that a significant number of images in the ImageNet dataset contain watermarks, making pre-trained networks susceptible to learning artifacts such as watermark patterns within their latent spaces. In this paper, we aim to assess the extent to which popular pre-trained architectures display such behavior and to determine which classes are most affected. Additionally, we examine the impact of watermarks on the extracted features. Contrary to the popular belief that the Chinese logographic watermarks impact the "carton" class only, our analysis reveals that a variety of ImageNet classes, such as "monitor", "broom", "apron" and "safe" rely on spurious correlations. Finally, we propose a simple approach to mitigate this issue in fine-tuned networks by ignoring the encodings from the feature-extractor layer of ImageNet pre-trained networks that are most susceptible to watermark imprints.