Simple Baselines for Image Restoration (NAFNet)
Optional · paper guide 1 of 4 for DLCV 01B · Building Blocks and How to Compose Them: Classification and Restoration Networks. Citation numbers refer to the reference list at the bottom of this page.
Navigate: this page and this track: Navigate: this page and this track
Challenge
Chen et al. [19] ask whether increasingly elaborate restoration blocks are necessary. They keep a plain single-stage U-Net and build a competitive block through controlled additions, then remove components that can be replaced by simpler gates. The result is a study of macro-architecture versus block design, not a claim that all nonlinear computation is unnecessary.
Method

Figure source: [19], PDF page 6. Original PDF.
Establish a useful baseline before simplifying. Starting from a convolutional network, the paper adds layer normalization, changes ReLU to GELU and adds channel attention. The stable baseline provides a reference against which a simplification can be evaluated. The normalization experiment also changes the learning rate, which matters when attributing its gain.
Replace a named activation with multiplication. After pointwise expansion and depthwise convolution, split an even-channel feature tensor into equal halves . SimpleGate is
denotes channel concatenation and elementwise multiplication. The output has half the input channels. This operation is nonlinear: scaling both branches by scales their product by . “Nonlinear Activation Free” refers to removing conventional activation functions such as GELU, not to turning the network into a linear filter.
Simplify channel attention too. For gated feature map , simplified channel attention has the form
averages spatial positions, is a learned channel projection, and the resulting channel vector broadcasts over space. Removing the original attention bottleneck/activation retains a data-dependent multiplicative channel gate.
A NAF block uses normalization, expansion, depthwise filtering, SimpleGate, SCA and projection in one residual path. A second normalized feed-forward path uses another SimpleGate and residual addition. Learnable residual scales stabilize the initial mapping. The encoder–decoder carries information across resolutions; the simple block is not a substitute for those skip and scale connections.
Read the figure and choose a baseline. Follow a feature tensor through the block: normalization prepares the input, spatial convolution mixes nearby evidence, and SimpleGate multiplies two channel groups before channel reweighting and the residual addition. The multiplication makes the mapping nonlinear even without a named activation. To judge a replacement block, keep the restoration task, width, training recipe and inference conversion fixed. Otherwise, an improvement cannot be assigned to removing activations. Use this as a compact learned baseline when paired training data are available; it does not estimate a physical blur kernel.
Results
The full denoising comparison in Table 6 reports 40.30 dB / 0.962 SSIM on SIDD with 65G MACs. The stronger baseline has the same displayed quality but 84G MACs. These full-model values should be kept separate from smaller ablation configurations.
For deblurring, Table 4’s final crop-training/test-time configuration reaches 33.69 dB / 0.967 SSIM on GoPro. Test-time local conversion is part of that setting; the block alone does not account for every improvement.
Ablations
Table 1 builds the baseline cumulatively. PlainNet at learning rate gives GoPro PSNR 28.51; the unstable run fails, while normalization permits the larger learning rate and reaches 31.90. That comparison changes both stability and optimization.
Table 2 gives baseline GoPro 32.35 dB, replacing GELU with SimpleGate 32.76, replacing attention with SCA 32.54, and both replacements 32.85. On SIDD the corresponding values are 39.85, 39.93, 39.95 and 39.96. Tasks differ in how much each change helps.
Table 5 reintroduces activations inside the gate. Some slightly improve SIDD while lowering GoPro performance, supporting a useful simple design without proving that one activation choice dominates all tasks. Table 3’s depth study also shows that more blocks increase latency and do not monotonically improve every metric.
Conclusion
NAFNet demonstrates that a carefully controlled plain architecture with multiplicative gates can match more elaborate restoration designs. Its argument depends on a strong baseline, explicit ablations and the distinction between named activation functions and nonlinear feature interactions.
Problems
The tested simplicity does not remove the need for training data, memory or global pooling. Crop-trained pooling statistics can differ at full-image inference, motivating the evaluated local conversion. A result on paired denoising/deblurring does not establish robustness to arbitrary real degradations, and equal operation counts need not imply equal device latency.
References
-
K. He, X. Zhang, S. Ren and J. Sun, “Deep Residual Learning for Image Recognition,” in Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016. CVF · arXiv
-
Z. Liu, H. Mao, C.-Y. Wu, C. Feichtenhofer, T. Darrell and S. Xie, “A ConvNet for the 2020s,” in Proc. IEEE/CVF CVPR, 2022. CVF · arXiv
-
A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer et al., “An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale,” in International Conference on Learning Representations (ICLR), 2021. arXiv
-
J. Liang, J. Cao, G. Sun, K. Zhang, L. Van Gool and R. Timofte, “SwinIR: Image Restoration Using Swin Transformer,” in Proc. IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), 2021. CVF · arXiv
-
T.-Y. Lin, P. Dollár, R. Girshick, K. He, B. Hariharan and S. Belongie, “Feature Pyramid Networks for Object Detection,” in Proc. IEEE CVPR, 2017. CVF · arXiv
-
Z. Liu, Y. Lin, Y. Cao, H. Hu, Y. Wei, Z. Zhang, S. Lin and B. Guo, “Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows,” in Proc. IEEE/CVF International Conference on Computer Vision (ICCV), 2021. CVF · arXiv
-
O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy et al., “ImageNet Large Scale Visual Recognition Challenge,” International Journal of Computer Vision, vol. 115, no. 3, pp. 211–252, 2015. doi
-
W. Yu, M. Luo, P. Zhou, C. Si, Y. Zhou, X. Wang, J. Feng and S. Yan, “MetaFormer Is Actually What You Need for Vision,” in Proc. IEEE/CVF CVPR, 2022. CVF · arXiv
-
A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, M. Andreetto and H. Adam, “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,” arXiv:1704.04861, 2017. arXiv
-
M. Sandler, A. Howard, M. Zhu, A. Zhmoginov and L.-C. Chen, “MobileNetV2: Inverted Residuals and Linear Bottlenecks,” in Proc. IEEE/CVF CVPR, 2018. CVF · arXiv
-
K. He, X. Zhang, S. Ren and J. Sun, “Identity Mappings in Deep Residual Networks,” in Proc. European Conference on Computer Vision (ECCV), LNCS, Springer, 2016. doi · arXiv
-
B. Lim, S. Son, H. Kim, S. Nah and K. M. Lee, “Enhanced Deep Residual Networks for Single Image Super-Resolution,” in Proc. IEEE CVPR Workshops (CVPRW), 2017. CVF · arXiv
-
H. Touvron, M. Cord, A. Sablayrolles, G. Synnaeve and H. Jégou, “Going Deeper With Image Transformers,” in Proc. IEEE/CVF ICCV, 2021. CVF
-
L. Chen, X. Chu, X. Zhang and J. Sun, NAFNet official implementation,
basicsr/models/archs/NAFNet_arch.py, GitHub (megvii-research/NAFNet), accessed 2026. code -
Y. Zhang, K. Li, K. Li, L. Wang, B. Zhong and Y. Fu, “Image Super-Resolution Using Very Deep Residual Channel Attention Networks,” in Proc. ECCV, 2018. CVF · arXiv
-
I. Radosavovic, R. P. Kosaraju, R. Girshick, K. He and P. Dollár, “Designing Network Design Spaces,” in Proc. IEEE/CVF CVPR, 2020. CVF · arXiv
-
J. Hu, L. Shen and G. Sun, “Squeeze-and-Excitation Networks,” in Proc. IEEE/CVF CVPR, 2018. CVF · arXiv
-
S. Woo, J. Park, J.-Y. Lee and I. S. Kweon, “CBAM: Convolutional Block Attention Module,” in Proc. ECCV, 2018. CVF · arXiv
-
L. Chen, X. Chu, X. Zhang and J. Sun, “Simple Baselines for Image Restoration,” in Proc. ECCV, LNCS, Springer, 2022. doi · arXiv
-
M.-H. Guo, T.-X. Xu, J.-J. Liu, Z.-N. Liu, P.-T. Jiang, T.-J. Mu, S.-H. Zhang, R. R. Martin, M.-M. Cheng and S.-M. Hu, “Attention Mechanisms in Computer Vision: A Survey,” Computational Visual Media, vol. 8, no. 3, pp. 331–368, 2022. doi · arXiv
-
S. W. Zamir, A. Arora, S. Khan, M. Hayat, F. S. Khan and M.-H. Yang, “Restormer: Efficient Transformer for High-Resolution Image Restoration,” in Proc. IEEE/CVF CVPR, 2022. CVF · arXiv
-
X. Chen, Z. Li, Y. Pu, Y. Liu, J. Zhou, Y. Qiao and C. Dong, “A Comparative Study of Image Restoration Networks for General Backbone Network Design,” in Proc. ECCV 2024, LNCS, Springer, 2025. doi · arXiv
-
S. Ioffe and C. Szegedy, “Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift,” in Proc. International Conference on Machine Learning (ICML), PMLR 37, 2015. PMLR
-
J. L. Ba, J. R. Kiros and G. E. Hinton, “Layer Normalization,” arXiv:1607.06450, 2016. arXiv
-
D. Ulyanov, A. Vedaldi and V. Lempitsky, “Instance Normalization: The Missing Ingredient for Fast Stylization,” arXiv:1607.08022, 2016. arXiv
-
Y. Wu and K. He, “Group Normalization,” in Proc. ECCV, 2018. CVF · arXiv
-
C. Ledig, L. Theis, F. Huszár, J. Caballero, A. Cunningham, A. Acosta, A. Aitken, A. Tejani, J. Totz, Z. Wang and W. Shi, “Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network,” in Proc. IEEE CVPR, 2017. CVF
-
X. Wang, K. Yu, S. Wu, J. Gu, Y. Liu, C. Dong, Y. Qiao and C. C. Loy, “ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks,” in Proc. ECCV Workshops, 2018. CVF · arXiv
-
K. Zhang, W. Zuo, Y. Chen, D. Meng and L. Zhang, “Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising,” IEEE Transactions on Image Processing, vol. 26, no. 7, pp. 3142–3155, 2017. doi · arXiv
-
Q. Huynh-Thu and M. Ghanbari, “Scope of Validity of PSNR in Image/Video Quality Assessment,” Electronics Letters, vol. 44, no. 13, pp. 800–801, 2008. doi
-
A. Abdelhamed, S. Lin and M. S. Brown, “A High-Quality Denoising Dataset for Smartphone Cameras,” in Proc. IEEE/CVF CVPR, 2018. CVF
-
S. Nah, T. H. Kim and K. M. Lee, “Deep Multi-Scale Convolutional Neural Network for Dynamic Scene Deblurring,” in Proc. IEEE CVPR, 2017. CVF
-
N. Shazeer, “GLU Variants Improve Transformer,” arXiv:2002.05202, 2020. arXiv
-
Y. N. Dauphin, A. Fan, M. Auli and D. Grangier, “Language Modeling with Gated Convolutional Networks,” in Proc. ICML, PMLR 70, 2017. PMLR
-
D. Hendrycks and K. Gimpel, “Gaussian Error Linear Units (GELUs),” arXiv:1606.08415, 2016. arXiv
-
W. Shi, J. Caballero, F. Huszár, J. Totz, A. P. Aitken, R. Bishop, D. Rueckert and Z. Wang, “Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network,” in Proc. IEEE CVPR, 2016. CVF · arXiv
-
A. Odena, V. Dumoulin and C. Olah, “Deconvolution and Checkerboard Artifacts,” Distill, vol. 1, no. 10, 2016. Distill · doi
-
A. Aitken, C. Ledig, L. Theis, J. Caballero, Z. Wang and W. Shi, “Checkerboard Artifact Free Sub-Pixel Convolution: A Note on Sub-Pixel Convolution, Resize Convolution and Convolution Resize,” arXiv:1707.02937, 2017. arXiv
-
O. Ronneberger, P. Fischer and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” in Proc. Medical Image Computing and Computer-Assisted Intervention (MICCAI), LNCS, Springer, pp. 234–241, 2015. doi · arXiv
-
G. Huang, Z. Liu, L. van der Maaten and K. Q. Weinberger, “Densely Connected Convolutional Networks,” in Proc. IEEE CVPR, 2017. CVF
-
L.-C. Chen, G. Papandreou, I. Kokkinos, K. Murphy and A. L. Yuille, “DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 40, no. 4, pp. 834–848, 2018. doi · arXiv
-
W. Luo, Y. Li, R. Urtasun and R. Zemel, “Understanding the Effective Receptive Field in Deep Convolutional Neural Networks,” in Advances in Neural Information Processing Systems (NeurIPS), 2016. NeurIPS · arXiv
-
J. Kim, J. K. Lee and K. M. Lee, “Accurate Image Super-Resolution Using Very Deep Convolutional Networks,” in Proc. IEEE CVPR, 2016. CVF · arXiv
-
H. Guo, J. Li, T. Dai, Z. Ouyang, X. Ren and S.-T. Xia, “MambaIR: A Simple Baseline for Image Restoration with State-Space Model,” in Proc. ECCV 2024, LNCS, Springer, 2025. doi · arXiv
-
ITU-T, “Recommendation T.81: Information technology – Digital compression and coding of continuous-tone still images – Requirements and guidelines” (the JPEG standard), 1992. ITU
-
T. Elsken, J. H. Metzen and F. Hutter, “Neural Architecture Search: A Survey,” Journal of Machine Learning Research, vol. 20, 2019. JMLR · arXiv
-
A. Khan, A. Sohail, U. Zahoora and A. S. Qureshi, “A Survey of the Recent Architectures of Deep Convolutional Neural Networks,” Artificial Intelligence Review, vol. 53, no. 8, pp. 5455–5516, 2020. doi · arXiv
-
S. Khan, M. Naseer, M. Hayat, S. W. Zamir, F. S. Khan and M. Shah, “Transformers in Vision: A Survey,” ACM Computing Surveys, vol. 54, no. 10s, 2022. doi · arXiv
-
S. Woo, S. Debnath, R. Hu, X. Chen, Z. Liu, I. S. Kweon and S. Xie, “ConvNeXt V2: Co-Designing and Scaling ConvNets With Masked Autoencoders,” in Proc. IEEE/CVF CVPR, 2023. CVF · arXiv
-
M. Goldblum, H. Souri, R. Ni, M. Shu, V. Prabhu, G. Somepalli, P. Chattopadhyay, M. Ibrahim et al., “Battle of the Backbones: A Large-Scale Comparison of Pretrained Models across Computer Vision Tasks,” in Advances in Neural Information Processing Systems (NeurIPS), 2023. arXiv
-
X. Chen, X. Wang, J. Zhou, Y. Qiao and C. Dong, “Activating More Pixels in Image Super-Resolution Transformer,” in Proc. IEEE/CVF CVPR, 2023. CVF · arXiv
-
H. Guo, Y. Guo, Y. Zha, Y. Zhang, W. Li, T. Dai, S.-T. Xia and Y. Li, “MambaIRv2: Attentive State Space Restoration,” in Proc. IEEE/CVF CVPR, 2025. CVF · arXiv
-
A. Araujo, W. Norris and J. Sim, “Computing Receptive Fields of Convolutional Neural Networks,” Distill, vol. 4, no. 11, 2019. Distill · doi
-
X. Ding, X. Zhang, N. Ma, J. Han, G. Ding and J. Sun, “RepVGG: Making VGG-style ConvNets Great Again,” in Proc. IEEE/CVF CVPR, pp. 13733–13742, 2021. CVF paper
-
S. Zagoruyko and N. Komodakis, “DiracNets: Training Very Deep Neural Networks Without Skip-Connections,” arXiv:1706.00388, 2017. arXiv
-
X. Ding, Y. Guo, G. Ding and J. Han, “ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks,” in Proc. IEEE/CVF ICCV, 2019. CVF
-
H. Zhao, J. Shi, X. Qi, X. Wang and J. Jia, “Pyramid Scene Parsing Network,” in Proc. IEEE CVPR, 2017. CVF
-
M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Benenson, U. Franke, S. Roth and B. Schiele, “The Cityscapes Dataset for Semantic Urban Scene Understanding,” in Proc. IEEE CVPR, 2016. CVF
-
E. D. Cubuk, B. Zoph, D. Mane, V. Vasudevan and Q. V. Le, “AutoAugment: Learning Augmentation Strategies From Data,” in Proc. IEEE/CVF CVPR, 2019. CVF
-
H. Zhang, M. Cisse, Y. N. Dauphin and D. Lopez-Paz, “mixup: Beyond Empirical Risk Minimization,” in Proc. ICLR, 2018. arXiv