ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks This paper presents ESRGAN, an enhanced version of the Super-Resolution Generative Adversarial Network (SRGAN), which aims to improve the visual quality of super-resolution (SR) results. ESRGAN addresses the issue of unpleasant artifacts often associated with SRGAN by refining three key components: network architecture, adversarial loss, and perceptual loss. Specifically, ESRGAN introduces the Residual-in-Residual Dense Block (RRDB) without batch normalization as the basic network unit, borrows the idea of relativistic GAN to improve the discriminator, and enhances the perceptual loss by using features before activation. These improvements result in ESRGAN achieving better visual quality with more realistic and natural textures, outperforming SRGAN in sharpness and details. ESRGAN won the first place in the PIRM2018-SR Challenge, demonstrating its superior performance in perceptual quality. The code for ESRGAN is available at <https://github.com/xintao/ESRGAN>.ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks This paper presents ESRGAN, an enhanced version of the Super-Resolution Generative Adversarial Network (SRGAN), which aims to improve the visual quality of super-resolution (SR) results. ESRGAN addresses the issue of unpleasant artifacts often associated with SRGAN by refining three key components: network architecture, adversarial loss, and perceptual loss. Specifically, ESRGAN introduces the Residual-in-Residual Dense Block (RRDB) without batch normalization as the basic network unit, borrows the idea of relativistic GAN to improve the discriminator, and enhances the perceptual loss by using features before activation. These improvements result in ESRGAN achieving better visual quality with more realistic and natural textures, outperforming SRGAN in sharpness and details. ESRGAN won the first place in the PIRM2018-SR Challenge, demonstrating its superior performance in perceptual quality. The code for ESRGAN is available at <https://github.com/xintao/ESRGAN>.