This article provides a comprehensive survey on recent advances in image super-resolution (SR) using deep learning techniques. The authors categorize existing SR methods into three main types: supervised SR, unsupervised SR, and domain-specific SR. They also cover important issues such as benchmark datasets and performance evaluation metrics. The survey highlights the contributions of deep learning to SR, including the development of various model frameworks, upsampling methods, network design strategies, and advanced convolution operations. The authors discuss the challenges and open issues in the field, emphasizing the need for further research in perception-distortion trade-offs and efficient SR models for smartphones. The survey concludes by identifying future directions and open issues, providing insights for the community.This article provides a comprehensive survey on recent advances in image super-resolution (SR) using deep learning techniques. The authors categorize existing SR methods into three main types: supervised SR, unsupervised SR, and domain-specific SR. They also cover important issues such as benchmark datasets and performance evaluation metrics. The survey highlights the contributions of deep learning to SR, including the development of various model frameworks, upsampling methods, network design strategies, and advanced convolution operations. The authors discuss the challenges and open issues in the field, emphasizing the need for further research in perception-distortion trade-offs and efficient SR models for smartphones. The survey concludes by identifying future directions and open issues, providing insights for the community.