• Home
  • Search
  • Transfer Learning for Text-To-Image Diffusion Models: Methods, Challenges, and Applications
  • https://doi.org/10.26562/ijirae.2025.v1211.12Copy DOI Icon

Transfer Learning for Text-To-Image Diffusion Models: Methods, Challenges, and Applications

  • Abstract
  • Literature Map
  • Similar Papers
Abstract

This survey paper presents a comprehensive analysis of transfer learning techniques in image synthesis using deep learning methods, with particular emphasis on text-to-image generation models. The field has witnessed remarkable progress through the evolution of Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and Diffusion Models. We systematically review state-of-the-art approaches including [1]StackGAN, [2]AttnGAN, DALL•E, Stable Diffusion, and their variants, examining their architectural innovations, training strategies, and performance characteristics. Our analysis covers critical aspects such as semantic alignment between text and images, multi-modal feature fusion, and cross-domain adaptation. We evaluate these methods across standard benchmarks including MS-COCO, CUB-200, and Oxford-102 Flowers datasets, discussing evaluation metrics like FID, IS, CLIP Score, and human evaluation protocols. Furthermore, we identify persistent challenges including computational complexity, semantic gap reduction, dataset limitations, and ethical considerations. The survey concludes with future research directions emphasizing lightweight architectures, multilingual support, improved controllability, and responsible AI deployment in image synthesis applications.

Similar Papers
  • PDF
  • Research Article
  • Citations6

GGADN: Guided generative adversarial dehazing network

  • Aug 03, 2021
  • Soft Computing
  • Jian Zhang +2
  • PDF
  • Research Article
  • Citations9

A Survey on Surface Defect Inspection Based on Generative Models in Manufacturing

  • Aug 02, 2024
  • Applied Sciences
  • Yu He +3
  • PDF
  • Research Article
  • Citations28

BEGAN v3: Avoiding Mode Collapse in GANs Using Variational Inference

  • Apr 23, 2020
  • Electronics
  • Sung-Wook Park +2
  • PDF
  • Research Article
  • Citations29

Generative Adversarial Networks for Creating Synthetic Nucleic Acid Sequences of Cat Genome.

  • Mar 28, 2022
  • International Journal of Molecular Sciences
  • Debapriya Hazra +2
  • Research Article
  • Citations1

SCGN: novel generative model using the convergence of latent space by training

  • Aug 01, 2020
  • Electronics Letters
  • H Kim +1
  • Conference Article
  • Citations3

Analysis of Generating Handwriting Based on GAN Model With Different Structures

  • Sep 24, 2021
  • Siqi Zheng
  • Research Article
  • Citations3

Scope of generative adversarial networks (GANs) in image processing

  • Jun 24, 2022
  • International journal of health sciences
  • Sudesh Kumari +1
  • PDF
  • Research Article
  • Citations9

Unveiling Hidden Insights in Gas Chromatography Data Analysis with Generative Adversarial Networks

  • Jul 07, 2024
  • Chemosensors
  • Namkyung Yoon +2
  • Research Article
  • Citations18

Deep learning based adaptive sequential data augmentation technique for the optical network traffic synthesis.

  • Jun 20, 2019
  • Optics Express
  • Jin Li +5
  • Research Article
  • Citations39

Towards automated 3D evaluation of water leakage on a tunnel face via improved GAN and self-attention DL model

  • Oct 04, 2023
  • Tunnelling and Underground Space Technology
  • Chen Wu +5
  • Research Article
  • Citations3

GAN-BASED SYNTHETIC MEDICAL IMAGE AUGMENTATION FOR CLASS IMBALANCED DERMOSCOPIC IMAGE ANALYSIS

  • Jan 01, 2025
  • Fractals
  • Amal Alshardan +5
  • Conference Article
  • Citations2

Evaluating Image Synthesis: A Modest Review of Techniques and Metrics

  • Sep 30, 2024
  • Roney Nogueira De Sousa +1
  • Research Article
  • Citations1

Exploring generative artificial intelligence: a comprehensive guide

  • Jan 06, 2026
  • PeerJ Computer Science
  • Rasha Shoitan +4
  • Research Article
  • Citations1

A Multidomain Generative Adversarial Network for Hoarse-to-Normal Voice Conversion

  • Oct 01, 2023
  • Journal of voice : official journal of the Voice Foundation
  • Minghang Chu +7
  • Conference Article

The Collapse of Data: Comparing the Resiliency of Different AI Models to Model Collapse Using AI-Generated Image Data

  • Dec 22, 2025
  • Kevin Audreylius +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.