STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Artificial Intelligence Research & Advances Review Article

A Comparison of Different Generative AI Models

  1. Arya Bandal Department of Electronics and Telecommunication, Rajgad Dnyanpeeth's Shri Chhatrapati Shivaji Maharaj College of Engineering, Savitribai Phule Pune University
  2. Samiya Attar Department of Electronics and Telecommunication, Rajgad Dnyanpeeth's Shri Chhatrapati Shivaji Maharaj College of Engineering, Savitribai Phule Pune University
  3. Sakshi Badak Department of Electronics and Telecommunication, Rajgad Dnyanpeeth's Shri Chhatrapati Shivaji Maharaj College of Engineering, Savitribai Phule Pune University

Abstract

Generative models have significantly advanced the field of artificial intelligence by allowing machines to produce complex and realistic outputs such as images, text, and other forms of data. Among the leading frameworks in this domain are generative adversarial networks (GANs), variational autoencoders (VAEs), and architectures based on Transformers. Each model offers specific benefits and drawbacks concerning design structure, training demands, and range of applications. This paper provides a detailed comparison of these generative techniques, with a focus on core aspects like accuracy, computational cost, and usability in real-world scenarios. We delve into the foundational concepts behind each model, evaluate their performance using widely accepted metrics, and analyze their effectiveness across tasks such as image generation, language modeling, and anomaly identification. The study outlines the trade-offs between flexibility, robustness, and scalability, aiming to guide practitioners in choosing the best-suited model for particular use cases. Finally, the paper discusses prospective research pathways to further enhance the power and versatility of generative models.

Keywords

References (11)

  1. Kingma DP, Welling M. Auto-encoding variational Bayes. [Preprint]. 2013. arXiv:1312.6114. doi:10.48550/arXiv.1312.6114.
  2. Saul LK, Weiss Y, Bottou L. Advances in Neural Information Processing Systems 17: Proceedings of the 2004 Conference. Cambridge (MA): MIT Press; 2005.
  3. Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, et al. Generative adversarial networks. Communications of the ACM. 2020;63(11):139-144. doi:10.1145/3422622
  4. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser Ł, Polosukhin I. Attention is all you need. In: von Luxburg U, Guyon I, Bengio S, Wallach H, Fergus R, Vishwanathan S, Garnett R, editors. Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS 2017). Red Hook (NY): Curran Associates Inc.; 2017. p. 5998–6008.
  5. Ahmad Z, Jaffri ZUA, Chen M, Bao S. Understanding GANs: fundamentals, variants, training challenges, applications, and open problems. Multimedia Tools and Applications. 2024;84(12):10347-10423. doi:10.1007/s11042-024-19361-y
  6. Wang J, Yang C, Xu Y, Shen Y, Li H, Zhou B. Improving GAN Equilibrium by Raising Spatial Awareness. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2022:11275-11283. doi:10.1109/cvpr52688.2022.01100
  7. Odaibo S. Tutorial: deriving the standard variational autoencoder (VAE) loss function. [Preprint]. 2019. arXiv:1907.08956. doi:10.48550/arXiv.1907.08956.
  8. Thrun S, Saul L, Schölkopf B. Advances in Neural Information Processing Systems 16: Proceedings of the 2003 Conference. Cambridge (MA): MIT Press; 2004. p.47–110.© STM Journals 2026. All Rights Reserved 22
  9. Margossian CC, Blei DM. Amortized variational inference: when and why? [Preprint]. 2023. arXiv:2307.11018. /doi:10.48550/arXiv.2307.11018.
  10. Zhang C, Butepage J, Kjellstrom H, Mandt S. Advances in Variational Inference. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2019;41(8):2008-2026. doi:10.1109/tpami.2018.2889774
  11. Hacker P, Engel A, Mauer M. Regulating ChatGPT and other Large Generative AI Models. 2023 ACM Conference on Fairness Accountability and Transparency. 2023:1112-1123. doi:10.1145/3593013.3594067
Support