Instance-wise Hard Negative Example Generation for Contrastive Learning in Unpaired Image-to-Image Translation

TL;DR

This paper introduces NEGCUT, a novel method that generates instance-wise hard negative examples online to enhance contrastive learning in unpaired image-to-image translation, leading to improved translation quality.

Contribution

The paper proposes a generator-based approach to produce hard negative examples tailored to each input, significantly boosting translation performance.

Findings

Achieves state-of-the-art results on three benchmark datasets.

Improves content preservation and translation quality.

Demonstrates effectiveness of instance-wise hard negatives.

Abstract

Contrastive learning shows great potential in unpaired image-to-image translation, but sometimes the translated results are in poor quality and the contents are not preserved consistently. In this paper, we uncover that the negative examples play a critical role in the performance of contrastive learning for image translation. The negative examples in previous methods are randomly sampled from the patches of different positions in the source image, which are not effective to push the positive examples close to the query examples. To address this issue, we present instance-wise hard Negative Example Generation for Contrastive learning in Unpaired image-to-image Translation (NEGCUT). Specifically, we train a generator to produce negative examples online. The generator is novel from two perspectives: 1) it is instance-wise which means that the generated examples are based on the input image, and 2) it can generate hard negative examples since it is trained with an adversarial loss. With the generator, the performance of unpaired image-to-image translation is significantly improved. Experiments on three benchmark datasets demonstrate that the proposed NEGCUT framework achieves state-of-the-art performance compared to previous methods.