Scholarlink

<p>Super-resolution (SR) is a crucial technique for enhancing image quality in various applications such as satellite imaging, surveillance, and medicine. This paper presents GANIRSR, a novel Generative Adversarial Network (GAN) incorporating Inception modules and Residual blocks for image super-resolution. The proposed model is designed to capture multiscale features and enhance feature learning, resulting in improved reconstruction accuracy. Trained on the Unsplash dataset and evaluated on SET14, GANIRSR achieves a Peak Signal-to-Noise Ratio (PSNR) of 29.23 dB, outperforming state-of-the-art methods. This study demonstrates the potential of integrating advanced GAN architectures for high-quality image super-resolution, with applications in diverse fields requiring enhanced image detail and clarity. The source code of the proposed model and the related datasets are available at <a href="https://github.com/Ashima-Jain2001/GAN-Inception-Residual-SuperResolution/tree/main">https://github.com/Ashima-Jain2001/GAN-Inception-Residual-SuperResolution/tree/main</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

GANIRSR: enhanced image super-resolution via generative adversarial network integrated with inception modules and residual blocks

  • 期刊论文

  • 发表日期：

    2025年12月27日

查看全文
  • Ashima Jain

摘要

Super-resolution (SR) is a crucial technique for enhancing image quality in various applications such as satellite imaging, surveillance, and medicine. This paper presents GANIRSR, a novel Generative Adversarial Network (GAN) incorporating Inception modules and Residual blocks for image super-resolution. The proposed model is designed to capture multiscale features and enhance feature learning, resulting in improved reconstruction accuracy. Trained on the Unsplash dataset and evaluated on SET14, GANIRSR achieves a Peak Signal-to-Noise Ratio (PSNR) of 29.23 dB, outperforming state-of-the-art methods. This study demonstrates the potential of integrating advanced GAN architectures for high-quality image super-resolution, with applications in diverse fields requiring enhanced image detail and clarity. The source code of the proposed model and the related datasets are available at https://github.com/Ashima-Jain2001/GAN-Inception-Residual-SuperResolution/tree/main.