Lightweight interactive feature inference network for single-image super-resolution
2024

Lightweight Interactive Feature Inference Network for Image Super-Resolution

publication Evidence: high

Author Information

Author(s): Wang Li, Li Xing, Tian Wei, Peng Jianhua, Chen Rui

Primary Institution: School of Computer and Software, Nanjing Vocational University of Industry Technology

Hypothesis

Can a lightweight interactive feature inference network improve image super-resolution performance while reducing computational costs?

Conclusion

The proposed interactive feature inference network (IFIN) achieves state-of-the-art image super-resolution accuracy with low computational demands.

Supporting Evidence

  • The IFIN network effectively captures both local and global image features.
  • Extensive experiments show that IFIN outperforms existing lightweight super-resolution methods.
  • IFIN maintains low computational costs while achieving high-quality image reconstruction.

Takeaway

This study created a new way to make blurry pictures clearer using a smart network that doesn't need a lot of computer power.

Methodology

The study developed a lightweight network that combines convolutional neural networks and transformers to enhance image super-resolution.

Limitations

The method has slower inference speeds compared to most CNN-based methods and uses a fixed upsampling strategy.

Digital Object Identifier (DOI)

10.1038/s41598-024-62633-8

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication