MPRNet: Multi-Path Residual Network for Lightweight Image Super Resolution

  title={MPRNet: Multi-Path Residual Network for Lightweight Image Super Resolution},
  author={Armin Mehri and Parichehr Behjati Ardakani and Angel Domingo Sappa},
  journal={2021 IEEE Winter Conference on Applications of Computer Vision (WACV)},
Lightweight super resolution networks have extremely importance for real-world applications. In recent years several SR deep learning approaches with outstanding achievement have been introduced by sacrificing memory and computational cost. To overcome this problem, a novel lightweight super resolution network is proposed, which improves the SOTA performance in lightweight SR and performs roughly similar to computationally expensive networks. Multi-Path Residual Network designs with a set of… 

Figures and Tables from this paper

Hierarchical Residual Attention Network for Single Image Super-Resolution

A new lightweight super-resolution model based on an efficient method for residual feature and attention aggregation that surpasses state-of-the-art performance in several datasets, while maintaining relatively low computation and memory footprint is introduced.

Thunder: Thumbnail based Fast Lightweight Image Denoising Network

Extensive experiments have been carried out on two real-world denoising benchmarks, demonstrating that the proposed Thunder outperforms the existing lightweight models and achieves competitive performance on PSNR and SSIM when compared with the complex designs.

Asymmetric Information Distillation Network for Lightweight Super Resolution

A large number of experiments show that the proposed asymmetric information distillation network (AIDN), designed to achieve image super resolution performance equivalent to SRResNet, achieves a better balance of performance and complexity than SOTA model.

MISSU: 3D Medical Image Segmentation via Self-distilling TransUNet

This work proposes to self-distill a Transformer-based UNet for medical image segmentation, which simultaneously learns global semantic information and local spatial-detailed features and achieves the best performance over previous state-of-the-art methods.

A Residual Learning Approach to Deblur and Generate High Frame Rate Video With an Event Camera

The deblurring task on traditional cameras directed by events to be a residual learning one is formulated, and corresponding network architectures for effective learning ofdeblurring and high frame rate video generation tasks are proposed.

QSAM-Net: Rain streak removal by quaternion neural network with self-attention module

A novel quaternion multi-stage multiscale neural network with a self-attention module called QSAM-Net to remove rain streaks is developed, which requires significantly fewer parameters by a factor of 3.98, over prior methods, while improving visual quality.

A multi-objective opposition-based barnacles mating optimization for image super resolution using hyper-Spectral images

This paper adopts a latest optimization algorithm called O-BMO with optimized Non-negative Structured Sparse Representation (NSSR), Adaptive Discrete Wavelet Transform (ADWT) and Optimized Deep Convolutional Neural Network for developing the enhanced image super-resolution model.

Robust Deep Ensemble Method for Real-world Image Denoising

This paper proposes a simple yet effective Bayesian deep ensemble (BDE) method for real-world image denoising, where several representative deep denoisers pre-trained with various training data settings can be fused to improve robustness forreal-world noisy images.

Denoising and Dehazing an Image in a Cascaded Pattern for Continuous Casting

By taking advantage of deep learning in a modeling complex formulation, this proposed algorithm, called Cascaded Denoising and Dehazing Net (CDDNet) reduces noise and hazy in a cascading pattern, and generalizes so well that processing a video from an operating continuous casting factory with CDDNet resulted in high visual quality.

NTIRE 2021 Challenge on Quality Enhancement of Compressed Video: Methods and Results

  • Ren YangR. Timofte Ru Wang
  • Computer Science
    2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
  • 2021
This paper reviews the first NTIRE challenge on quality enhancement of compressed video, with a focus on the proposed methods and results, and gauge the state of the art of video quality enhancement.



Fast, Accurate, and, Lightweight Super-Resolution with Cascading Residual Network

This paper designs an architecture that implements a cascading mechanism upon a residual network that achieves performance comparable to that of state-of-the-art methods and presents variant models of the proposed cascading residual network to further improve efficiency.

Multi-attention Based Ultra Lightweight Image Super-Resolution

MAFFSRN consists of proposed feature fusion groups that serve as a feature extraction block and the MAB with a cost-efficient attention mechanism (CEA) helps to refine and extract the features using multiple attention mechanisms.

Ultra Lightweight Image Super-Resolution with Multi-Attention Layers

The comprehensive experiments show the superiority of the proposed Multi-Attentive Feature Fusion Super-Resolution Network over the existing state-ofthe-art in memory usage, floating-point operations (FLOPs) and number of parameters, respectively.

Enhanced Deep Residual Networks for Single Image Super-Resolution

This paper develops an enhanced deep super-resolution network (EDSR) with performance exceeding those of current state-of-the-art SR methods, and proposes a new multi-scale deepsuper-resolution system (MDSR) and training method, which can reconstruct high-resolution images of different upscaling factors in a single model.

Second-Order Attention Network for Single Image Super-Resolution

Experimental results demonstrate the superiority of the SAN network over state-of-the-art SISR methods in terms of both quantitative metrics and visual quality.

Residual Feature Aggregation Network for Image Super-Resolution

This work proposes a novel residual feature aggregation (RFA) framework for more efficient feature extraction and proposes an enhanced spatial attention (ESA) block to make the residual features to be more focused on critical spatial contents.

Image Super-Resolution Using Very Deep Residual Channel Attention Networks

This work proposes a residual in residual (RIR) structure to form very deep network, which consists of several residual groups with long skip connections, and proposes a channel attention mechanism to adaptively rescale channel-wise features by considering interdependencies among channels.

Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution

This paper proposes the Laplacian Pyramid Super-Resolution Network (LapSRN) to progressively reconstruct the sub-band residuals of high-resolution images and generates multi-scale predictions in one feed-forward pass through the progressive reconstruction, thereby facilitates resource-aware applications.

Fast, Accurate and Lightweight Super-Resolution with Neural Architecture Search

This work handles super-resolution with a multi-objective approach, and proposes an elastic search tactic at both micro and macro level, based on a hybrid controller that profits from evolutionary computation and reinforcement learning.

Efficient Single Image Super-Resolution via Hybrid Residual Feature Learning with Compact Back-Projection Network

  • Feiyang ZhuQijun Zhao
  • Computer Science
    2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)
  • 2019
This paper proposes an efficient SISR method via learning hybrid residual features, based on which the residual HR image can be reconstructed, and proposes a compact back-projection network that can simultaneously generate features in both LR and HR space by cascading up-and down-sampling layers with small-sized filters.