HINet: Half Instance Normalization Network for Image Restoration

By Liangyu Chen, Xin Lu, J. Zhang, Xiaojie Chu, Chengpeng Chen. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2021

683 citations recorded by the source corpus.

In this paper, we explore the role of Instance Normalization in low-level vision tasks. Specifically, we present a novel block: Half Instance Normalization Block (HIN Block), to boost the performance of image restoration networks. Based on HIN Block, we design a simple and powerful multi-stage network named HINet, which consists of two subnetworks. With the help of HIN Block, HINet surpasses the state-of-the-art (SOTA) on various image restoration tasks. For image denoising, we exceed it 0.11dB and 0.28 dB in PSNR on SIDD dataset, with only 7.5% and 30% of its multiplier-accumulator operations (MACs), 6.8× and 2.9× speedup respectively. For image deblurring, we get comparable performance with 22.5% of its MACs and 3.3× speedup on REDS and GoPro datasets. For image deraining, we exceed it by 0.3 dB in PSNR on the average result of multiple datasets with 1.4× speedup. With HINet, we won the 1st place on the NTIRE 2021 Image Deblurring Challenge - Track2. JPEG Artifacts, with a PSNR of 29.70.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/9b8cbe9a636b00b3543d8e1f98c24de12d6c3d17

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

MAXIM: Multi-Axis MLP for Image Processing (2022, methodology): “On denoising, for example, MAXIM-3S has only 21% FLOPs and 70% parameters of MIRNet [107]; on deblurring, our MAXIM-3S model requires only 25% of the number of parameters of the previous best model HINet [15], and merely 19% of the number of parameters of the Transformer model IPT [14].”
Deblurring via Stochastic Refinement (2021, methodology): “Another promising direction is to replace our initial predictor and de-noiser network with U-Net architectures that are optimized for both distortion and run time [11,14,75]. difference is size, where the starting channel count is 64 for the initial predictor and 32 for the denoiser.”
Simple Baselines for Image Restoration (2022, methodology): “Inter-block Complexity [35,4] are multi-stage networks, i.”
MST++: Multi-stage Spectral-wise Transformer for Efficient Spectral Reconstruction (2022, methodology): “We also train MIRNet [84], MPRNet [85], Restormer [83], HINet [21], and MST [13] families.”