DeblurGAN is an end-to-end learned method for motion deblurring, utilizing a conditional GAN and content loss. It achieves state-of-the-art performance in both structural similarity and visual appearance, outperforming the closest competitor, DeepDeblur, by a factor of five in speed. The method introduces a novel approach to generate synthetic motion-blurred images from sharp ones, enhancing dataset augmentation. DeblurGAN's performance is evaluated through object detection on blurred images, demonstrating its effectiveness in improving detection accuracy. The model, code, and dataset are available online. The paper also discusses related work in image deblurring and generative adversarial networks, and presents a detailed architecture and training process for DeblurGAN.DeblurGAN is an end-to-end learned method for motion deblurring, utilizing a conditional GAN and content loss. It achieves state-of-the-art performance in both structural similarity and visual appearance, outperforming the closest competitor, DeepDeblur, by a factor of five in speed. The method introduces a novel approach to generate synthetic motion-blurred images from sharp ones, enhancing dataset augmentation. DeblurGAN's performance is evaluated through object detection on blurred images, demonstrating its effectiveness in improving detection accuracy. The model, code, and dataset are available online. The paper also discusses related work in image deblurring and generative adversarial networks, and presents a detailed architecture and training process for DeblurGAN.