Enhancing Robustness and Efficiency in GAN-Based Single Image Super-Resolution

This presentation explores challenges in GAN-based Super-Resolution Image Synthesis (SISR) models and proposes architectural improvements benchmarked against state-of-the-art methods. It begins with an introduction to SISR applications and challenges, followed by a review of deep learning advancements and research gaps. Leveraging the Research Onion framework, the methodology outlines model development and evaluation processes. Key evaluation metrics and expected contributions, including a...