Advancing Text-to-Image Generation: A Comparative Study of StyleGAN-T and Stable Diffusion 3 under Neutrosophic Sets
Recent advances in generative models have revolutionized the technology employed for image synthesis quite significantly, and two paradigms—GANs and diffusion-based models—are leading the pack of innovation. This paper outlines an extensive comparison and analysis of some of the best models across b...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of New Mexico
2025-06-01
|
| Series: | Neutrosophic Sets and Systems |
| Subjects: | |
| Online Access: | https://fs.unm.edu/NSS/44TextImage.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Recent advances in generative models have revolutionized the technology employed for image synthesis quite significantly, and two paradigms—GANs and diffusion-based models—are leading the pack of innovation. This paper outlines an extensive comparison and analysis of some of the best models across both paradigms, namely StyleGAN-T, DF-GAN, AttnGAN, and BigGAN on the GAN side and Stable Diffusion 3 (SD3), DALL·E 3, Midjourney v6, and Imagen 2 on the diffusion side. We systematically inspect the architectural design, training protocols, text-conditioning processes, and domain adaptability of each model, highlighting how they address text-to-image generation challenges differently. Through qualitative and quantitative measurements—such as FID, CLIP Score, human preference surveys, and compositional accuracy, the work reveals performance tradeoffs concerning speed, control, creativity, semantic alignment, and photorealism. We use the Neutrosophic Set model to select the best model based on these evaluation matrices. We have different scores for each model based on evaluation matrices. So, the neutrosophic set is used to overcome the uncertainty information. We use the COPRAS method to rank the models and select the best one based on the evaluation matrix weights. |
|---|---|
| ISSN: | 2331-6055 2331-608X |