PRRN

Progressive Representation Recalibration for Lightweight Super-resolution [pdf]

Ruimian Wen, Zhijing Yang, Tianshui Chen, Hao Li, Kai Li

Guangdong University of Technology, ZEGO

NTIRE 2022 Workshop and Challenge @ CVPR 2022

Our Methods PRRN also participated in Efficient Super-Resolution Challenge. [Challenge Report]

---

Baseline model (IMDN)

• Number of parameters: 893,936 (0.89M)

  number_parameters = sum(map(lambda x: x.numel(), model.parameters()))

• Average PSNR on validation data: 29.13 dB

• Average inference time (Titan Xp) on validation data: 0.10 second

  Note: The best average inference time among three trials is selected.

Run test_demo.py to test the model

Our model (PRRN)

• Number of parameters: 414008

• Average PSNR on validation data: 29.05 dB

• Average inference time (Tesla V100) on validation data: 0.05 second

How to use the code during test phase.

1. git clone https://github.com/house-leo/PRRN
2. Put your model script under the models folder.
3. Put your pretrained model under the model_zoo folder.
4. Modify model_path in test_demo.py. Modify the imported models.
5. python test_demo.py

Citation:

If you find this work useful for your research, please cite:

@article{WEN2022240,
title = {Progressive representation recalibration for lightweight super-resolution},
journal = {Neurocomputing},
author = {Ruimian Wen and Zhijing Yang and Tianshui Chen and Hao Li and Kai Li},
volume = {504},
pages = {240-250},
year = {2022},
issn = {0925-2312}
}