Papers › Improving Text-to-Image Synthesis Using Contrastive Learning
Improving Text-to-Image Synthesis Using Contrastive Learning
Hui Ye, Xiulong Yang, Martin Takac, Rajshekhar Sunderraman, Shihao Ji
The goal of text-to-image synthesis is to generate a visually realistic image that matches a given text description. In practice, the captions annotated by humans for the same image have large variance in terms of contents and the choice of words. The linguistic discrepancy between the captions of the identical image leads to the synthetic images deviating from the ground truth. To address this issue, we propose a contrastive learning approach to improve the quality and enhance the semantic consistency of synthetic images. In the pretraining stage, we utilize the contrastive learning approach to learn the consistent textual representations for the captions corresponding to the same image. Furthermore, in the following stage of GAN training, we employ the contrastive learning method to enhance the consistency between the generated images from the captions related to the same image. We evaluate our approach over two popular text-to-image synthesis models, AttnGAN and DM-GAN, on datasets CUB and COCO, respectively. Experimental results have shown that our approach can effectively improve the quality of synthetic images in terms of three metrics: IS, FID and R-precision. Especially, on the challenging COCO dataset, our approach boosts the FID signifcantly by 29.60% over AttnGAN and by 21.96% over DM-GAN.
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Text-to-Image Generation | COCO (Common Objects in Context) | DM-GAN+CL | FID | 20.79 | #44 of 69 | Archive leaderboard | report |
| Text-to-Image Generation | COCO (Common Objects in Context) | DM-GAN+CL | Inception score | 33.34 | #44 of 69 | Archive leaderboard | report |
| Text-to-Image Generation | COCO (Common Objects in Context) | AttnGAN+CL | FID | 23.93 | #48 of 69 | Archive leaderboard | report |
| Text-to-Image Generation | COCO (Common Objects in Context) | AttnGAN+CL | Inception score | 25.70 | #48 of 69 | Archive leaderboard | report |
| Text-to-Image Generation | CUB | DM-GAN+CL | FID | 14.38 | #10 of 20 | Archive leaderboard | report |
| Text-to-Image Generation | CUB | DM-GAN+CL | Inception score | 4.77 | #10 of 20 | Archive leaderboard | report |
| Text-to-Image Generation | CUB | AttnGAN+CL | FID | 16.34 | #12 of 20 | Archive leaderboard | report |
| Text-to-Image Generation | CUB | AttnGAN+CL | Inception score | 4.42 | #12 of 20 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections