Pradeep Kumar

Dr. Pradeep Kumar


Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications



26 February 2025
Verified
Book Chapter
Closed access
PDF file
JVJayavrinda Vrindavanam
Jayavrinda Vrindavanam
PKPradeep Kumar
Pradeep Kumar
ABAbhilash S. Bharadwaj
Abhilash S. Bharadwaj
GKGaurav Kamath
Gaurav Kamath
NCN. Chandrashekar
N. Chandrashekar
ANAnirudh Narayanan
Anirudh Narayanan

2 January 2026
Verified
Book Chapter
Open access
PDF file
PKPradeep Kumar
Pradeep Kumar
SISalma Itagi
Salma Itagi
NVN. Vinutha
N. Vinutha
TMT. L. Manjushree
T. L. Manjushree
MAMonika Agarwal
Monika Agarwal
JVJayavrinda Vrindavanam
Jayavrinda Vrindavanam
This work explores the application of Enhanced Super Resolution Generative Adversarial Networks (ESRGAN) basically used for image resolution enhancement. Due to viable requirement for high-resolution images has increased across various domains, including medical imaging, surveil-lance systems, and digital content creation. ESRGAN has emerged as a prominent deep learning tool which can generate high-quality, visually appealing images from low-resolution inputs. This work provides an in-depth analysis of the underlying principles and architecture of ESRGAN. It discusses the generator and discriminator networks, as well as the adversarial training process and perceptual loss function. By understanding the inner workings of ESRGAN. To estimate the performance of ESRGAN, comprehensive experimentations were conducted. A suitable dataset was selected and preprocessed, and ESRGAN was trained using appropriate training procedures. Various evaluation metrics were employed to assess the image resolution produced. Existing models were assessed as well for evaluating the advantages of ESRGAN over other super-resolution approaches. The results demonstrate the effectiveness of ESRGAN in significantly enhancing image resolution. The quantitative evaluation reveals best performance in terms of metrics such as peak signal to noise ratio (PSNR) the structural similarity index (SSIM). Moreover, the qualitative evaluation demonstrates visually appealing results, with sharper edges, finer textures, and increased overall image quality. Despite its successes, ESRGAN has certain limitations and challenges. The paper discusses these limitations and suggests potential avenues for future research, including improving ESRGAN performance, addressing artifacts or distortions in the generated images, and exploring novel architectures and techniques. ESRGAN presents a promising approach to image super-resolution, offering significant enhancements in image quality and details. Its applications span various domains, making it a valuable tool for researchers, practitioners, and industries reliant on high-resolution image-ry. This paper highlights the use of ESRGAN, its evaluation, and its potential future advancements.