Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Generative Artificial Intelligence: A Systematic Review and Applications

Sandeep Singh Sengar, Affan Bin Hasan, Sanjay Kumar, Fiona Carroll

TL;DR

This paper conducts a systematic review of Generative AI, covering core architectures (GANs, Transformers, VAEs, Diffusion) from 2012 to 2023 and their progress toward real-world tasks. It analyzes advancements across image translation, video synthesis, NLP, and knowledge graph generation, detailing datasets, evaluation metrics, and application-specific models. The authors also address ethical, safety, and interpretability considerations, arguing for responsible AI development and governance. Overall, the work highlights the growing practical impact of GenAI and outlines trajectories for architecture refinement, cross-domain applications, and human–AI collaboration.

Abstract

In recent years, the study of artificial intelligence (AI) has undergone a paradigm shift. This has been propelled by the groundbreaking capabilities of generative models both in supervised and unsupervised learning scenarios. Generative AI has shown state-of-the-art performance in solving perplexing real-world conundrums in fields such as image translation, medical diagnostics, textual imagery fusion, natural language processing, and beyond. This paper documents the systematic review and analysis of recent advancements and techniques in Generative AI with a detailed discussion of their applications including application-specific models. Indeed, the major impact that generative AI has made to date, has been in language generation with the development of large language models, in the field of image translation and several other interdisciplinary applications of generative AI. Moreover, the primary contribution of this paper lies in its coherent synthesis of the latest advancements in these areas, seamlessly weaving together contemporary breakthroughs in the field. Particularly, how it shares an exploration of the future trajectory for generative AI. In conclusion, the paper ends with a discussion of Responsible AI principles, and the necessary ethical considerations for the sustainability and growth of these generative models.

Generative Artificial Intelligence: A Systematic Review and Applications

TL;DR

This paper conducts a systematic review of Generative AI, covering core architectures (GANs, Transformers, VAEs, Diffusion) from 2012 to 2023 and their progress toward real-world tasks. It analyzes advancements across image translation, video synthesis, NLP, and knowledge graph generation, detailing datasets, evaluation metrics, and application-specific models. The authors also address ethical, safety, and interpretability considerations, arguing for responsible AI development and governance. Overall, the work highlights the growing practical impact of GenAI and outlines trajectories for architecture refinement, cross-domain applications, and human–AI collaboration.

Abstract

In recent years, the study of artificial intelligence (AI) has undergone a paradigm shift. This has been propelled by the groundbreaking capabilities of generative models both in supervised and unsupervised learning scenarios. Generative AI has shown state-of-the-art performance in solving perplexing real-world conundrums in fields such as image translation, medical diagnostics, textual imagery fusion, natural language processing, and beyond. This paper documents the systematic review and analysis of recent advancements and techniques in Generative AI with a detailed discussion of their applications including application-specific models. Indeed, the major impact that generative AI has made to date, has been in language generation with the development of large language models, in the field of image translation and several other interdisciplinary applications of generative AI. Moreover, the primary contribution of this paper lies in its coherent synthesis of the latest advancements in these areas, seamlessly weaving together contemporary breakthroughs in the field. Particularly, how it shares an exploration of the future trajectory for generative AI. In conclusion, the paper ends with a discussion of Responsible AI principles, and the necessary ethical considerations for the sustainability and growth of these generative models.
Paper Structure (21 sections, 1 equation, 9 figures, 8 tables)

This paper contains 21 sections, 1 equation, 9 figures, 8 tables.

Table of Contents

  1. Introduction
  2. What is Generative Artificial Intelligence?
  3. Generative Adversarial Networks
  4. Transformers
  5. Variational Autoencoders
  6. Diffusion Models
  7. Evolution of Generative AI Models: A Look at Earlier Variants
  8. Earlier GAN Variants
  9. Earlier Transformer Variants
  10. Earlier Variational Autoencoder Variants
  11. Earlier Diffusion Model Variants
  12. Advancements in Generative AI and Their Diverse Applications
  13. Generative AI for Image Translation
  14. Generative AI for Video Synthesis and Generation
  15. Generative AI for Natural Language Processing
  16. ...and 6 more sections

Figures (9)

  • Figure 1: GAN- Generator and discriminator working
  • Figure 2: Basic GANS Variants
  • Figure 3: qualitative outcomes obtained from various translation techniques employed to generate T2 images from T1 images within the unpaired BraTs2018 dataset, Source: yan2022swin
  • Figure 4: unpaired UVCGAN vs Others image-to-image translation, Source: torbunov2023uvcgan
  • Figure 5: Highlighted with red boxes and magnified for emphasis, the images are presented in the following order: (a) SAR images, (b) Pix2Pix, (c) CycleGAN, (d) S-cycle-GAN, (e) NICE-GAN, (f) GANILLA, (g) CFRWD-GAN, and (h) ground truth optical images., Source: wei2023cfrwd
  • ...and 4 more figures