Lightweight Real-Time Image Super-Resolution Network for 4K Images

Ganzorig Gankuyag*, Kihwan Yoon*, Jinman Park, Haeng Seon Son, Kyoungwon Min,
Korea Electronics Technology Institute
*equal contribution

Abstract

Single-image super-resolution technology has become a topic of extensive research in various applications, aiming to enhance the quality and resolution of degraded images obtained from low-resolution sensors. However, most existing studies on single-image super-resolution have primarily focused on developing deep learning networks operating on high-performance graphics processing units.

Therefore, this study proposes a lightweight real-time image super-resolution network for 4K images. Furthermore, we applied a reparameterization method to improve the network performance without incurring additional computational costs. The experimental results demonstrate that the proposed network achieves a PSNR of 30.15 dB and an inference time of 4.75 ms on an RTX 3090Ti device, as evaluated on the NTIRE 2023 Real-Time Super-Resolution validation scale X3 dataset. The code is available at https://github.com/Ganzooo/LRSRN.

Acknowledgement

This work was supported by Institute of Information \& communications Technology Planning \& Evaluation (IITP) grant funded by the Korea government(MSIT) (No.2021-0-00800, Development of Driving Environment Data Transformation and Data Verification Technology for the Mutual Utilization of Self-driving Learning Data for Different Vehicles)

BibTeX


      @inproceedings{gankhuyag2023lightweight,
        title={Lightweight real-time image super-resolution network for 4k images},
        author={Gankhuyag, Ganzorig and Yoon, Kihwan and Park, Jinman and Son, Haeng Seon and Min, Kyoungwon},
        booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
        pages={1746--1755},
        year={2023}
      }