Dual Contrastive Loss and Attention for GANs

By Ning Yu, Guilin Liu, Aysegul Dundar, Andrew Tao, Bryan Catanzaro, Larry S. Davis, Mario Fritz. 2021

70 citations recorded by the source corpus.

Generative Adversarial Networks (GANs) produce impressive results on unconditional image generation when powered with large-scale image datasets. Yet generated images are still easy to spot especially on datasets with high variance (e.g. bedroom, church). In this paper, we propose various improvements to further push the boundaries in image generation. Specifically, we propose a novel dual contrastive loss and show that, with this loss, discriminator learns more generalized and distinguishable representations to incentivize generation. In addition, we revisit attention and extensively experiment with different attention blocks in the generator. We find attention to be still an important module for successful image generation even though it was not used in the recent state-of-the-art models. Lastly, we study different attention architectures in the discriminator, and propose a reference attention mechanism. By combining the strengths of these remedies, we improve the compelling state-of-the-art Fréchet Inception Distance (FID) by at least 17.5% on several benchmark datasets. We obtain even more significant improvements on compositional synthetic scenes (up to 47.5% in FID).

Subfields: Computer Vision and Pattern Recognition, Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Bolei Zhou
2. Yujun Shen
3. Ceyuan Yang
4. Yinghao Xu
5. Luc Van Gool
6. Mario Fritz
7. Ming-Hsuan Yang
8. Chen Change Loy

In Generative Models

1. Yujun Shen
2. Bolei Zhou
3. Yinghao Xu
4. Ceyuan Yang
5. Mario Fritz
6. Ning Yu
7. Daniel Cohen-Or
8. Sahar Abdelnabi

Citations in the corpus

• Data-Efficient Instance Generation from Instance Discrimination (2021, background): “Akin to supervised contrastive loss [27], some concurrent work [20, 21, 43] reformulates the conventional bi-classification task (i.”
• Artificial Fingerprinting for Generative Models: Rooting Deepfake Attribution in Training Data (2020, background): “In the past years, photorealistic image generation has been rapidly evolving, benefiting from the invention of generative adversarial networks (GANs) [16] and its successive breakthroughs [39, 17, 35, 5, 25, 26, 27, 50, 51].”