README.md

June 13, 2025 · View on GitHub

Chasing Faster ConvNet for Efficient Super-Resolution

Overview: The repository records a path of chasing faster ConvNet.

The repo is still under construction!

☁️ EFDN for NTIRE 2022 ESR

Edge-enhanced Feature Distillation Network for Efficient Super-Resolution
Yan Wang
Nankai University

Summary: 5th solution of Model Complexity in the NTIRE 2022 Challenge on Efficient Super-Resolution. Involoving the modification of convolution and network architecture.

🌟 Convolution: edge-ehanced reparameter block (EDBB) with a corresponding edge loss .
📦 Attention: original ESA.
📦 Backbone: backbone searched by network-level NAS.

🌥️ PFDN for NTIRE 2023 ESR

Partial Feature Distillation Network for Efficient Super-Resolution
Yan Wang, Erlin Pan^†, Qixuan Cai, Xinan Dai
Nankai University, University of Electronic Science and Technology of China, Tianjin University

Summary: Winner of Overall Evaluation and 4th of Runtime in the NTIRE 2023 Challenge on Efficient Super-Resolution. Involoving the modification of convolution and network architecture.

⭐️ Convolution: integrating partial convolution and RRRB.
📦 Attention: efficient ESA.
📦 Backbone: ResNet-style backbone.

_Model	_Runtime[ms]	_Params[M]	_Flops[G]	_Acts[M]	_{GPU Mem[M]}
RFDN	35.54	0.433	27.10	112.03	788.13
PFDN	20.49	0.272	16.76	65.10	296.45

⛅️ PFDNLite for NTIRE 2024 ESR

Lightening Partial Feature Distillation Network for Efficient Super-Resolution
Yan Wang, Yi Liu, Qing Wang, Gang Zhang, Liou Zhang, Shijie Zhao^†
Nankai University, ByteDance

Summary: 3rd of Overall Evaluation and 3rd of Runtime in the NTIRE 2024 Challenge on Efficient Super-Resolution. Involoving the modification of convolution, attention and network pruning.

📦 Convolution: RepMBConv in PlainUSR.
📦 Attention: LIA in PlainUSR.
⭐️ Backbone: ABPN-style backbone and block pruning.

🌤️ PlainUSR for ACCV 2024

PlainUSR: Chasing Faster ConvNet for Efficient Super-Resolution
Yan Wang, Yusen Li^†, Gang Wang, Xiaoguang Liu
Nankai University

Summary: we present PlainUSR incorporating three pertinent modifications (convolution, attention, and backbone) to expedite ConvNet for efficient SR.

🌟 Convolution: Reparameterized MobileNetV3 Convolution (RepMBConv).
⭐️ Attention: Local Importance-based Attention (LIA).
🌟 Backbone: Plain U-Net.

🌤️ ESPAN for NTIRE 2025 ESR

Expanded SPAN for Efficient Super-Resolution
Qing Wang, Yan Wang, Hongyu An, Yi Liu, Liou Zhang, Shijie Zhao^†
ByteDance

Summary: 5th of Overall Evaluation and 3rd of Runtime in the NTIRE 2024 Challenge on Efficient Super-Resolution.

@inproceedings{wang2022edge,
  title={Edge-enhanced Feature Distillation Network for Efficient Super-Resolution},
  author={Wang, Yan},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
  pages={777--785},
  year={2022}
}

@inproceedings{wang2024plainusr,
  title={PlainUSR: Chasing Faster ConvNet for Efficient Super-Resolution},
  author={Wang, Yan and Li, Yusen and Wang, Gang and Liu, Xiaoguang},
  booktitle={Proceedings of the Asian Conference on Computer Vision},
  pages={4262--4279},
  year={2024}
}

@inproceedings{wang2025expanded,
  title={Expanded SPAN for Efficient Super-Resolution},
  author={Wang, Qing and Wang, Yang and An, Hongyu and Liu, Yi and Zhang, Liou and Zhao, Shijie},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
  pages={967--976},
  year={2025}
}