Authors : Shyla N; Himanshu Negi; Aditya J Shetty; Abhimanyu Singh Kushwah; Sudhiti Khar

Volume/Issue : Volume 9 - 2024, Issue 11 - November

---

Google Scholar : https://tinyurl.com/yckdn4h3

DOI : https://doi.org/10.38124/ijisrt/24nov1693

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

---

Abstract : Generative Adversarial Networks (GANs) have revolutionized computer vision, enabling tasks such as realistic face generation, image super-resolution, and synthetic data creation. This survey explores various GAN models and methodologies with a focus on face generation. Special emphasis is placed on advancements in stabilizing GAN training, mitigating mode col- lapse, utilizing synthetic data for face recognition, and enhancing the robustness of GANs against adversarial attacks.

Keywords : Generative Adversarial Networks, Face Generation, Mode Collapse, Synthetic Data, Adversarial Robustness, Computer Vision.

References :

1. Generative Adversarial Networks for Image Super-Resolution
2. DCGAN-based Face Generation for Low-Resource Domains
3. Improving GAN Training Stability via Wasserstein GAN and Gradient Penalty
4. Conditional GANs for Controlled Face Synthesis
5. Exploring Mode Collapse in GANs and Techniques to Mitigate It
6. Face Recognition with Synthetic Data: A GAN-based Approach
7. Adversarial Attacks and Defenses in GAN-based Face Generation