mlsanityData to 7 Oct 2026

RealViformer: Investigating Attention for Real-World Video Super-Resolution

arXiv 2407.13987 ↗First posted 19 Jul 2024cs.CVcited by 33

Not on the map. Why?

Abstract (arXiv)

In real-world video super-resolution (VSR), videos suffer from in-the-wild degradations and artifacts. VSR methods, especially recurrent ones, tend to propagate artifacts over time in the real-world setting and are more vulnerable than image super-resolution. This paper investigates the influence of artifacts on commonly used covariance-based attention mechanisms in VSR. Comparing the widely-used spatial attention, which computes covariance over space, versus the channel attention, we observe that the latter is less sensitive to artifacts. However, channel attention leads to feature redundancy, as evidenced by the higher covariance among output channels. As such, we explore simple techniques such as the squeeze-excite mechanism and covariance-based rescaling to counter the effects of high channel covariance. Based on our findings, we propose RealViformer. This channel-attention-based real-world VSR framework surpasses state-of-the-art on two real-world VSR datasets with fewer parameters and faster runtimes. The source code is available at https://github.com/Yuehan717/RealViformer.

Key facts. Cited by 33 papers in the map (data through 7 Oct 2026). Builds on 34 of its 43 references that are in the map. First posted on arXiv 19 Jul 2024.

Builds on

34 of 43 references.

References · 34

  1. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  2. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  3. 2022Generative Adversarial Networkscited by 5,603
  4. 2017Squeeze-and-Excitation Networkscited by 4,763
  5. 2016Perceptual Losses for Real-Time Style Transfer and Super-Resolutioncited by 2,678
  6. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,869
  7. 2021Restormer: Efficient Transformer for High-Resolution Image Restorationcited by 1,618
  8. 2018Image Super-Resolution Using Very Deep Residual Channel Attention Networkscited by 1,256
  9. 2021VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learningcited by 1,013
  10. 2021Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Datacited by 961
  11. 2020Pre-Trained Image Processing Transformercited by 891
  12. 2021Designing a Practical Degradation Model for Deep Blind Image Super-Resolutioncited by 524
22 more
  1. 2019EDVR: Video Restoration with Enhanced Deformable Convolutional Networkscited by 492
  2. 2022Activating More Pixels in Image Super-Resolution Transformercited by 433
  3. 2016Optical Flow Estimation using a Spatial Pyramid Networkcited by 421
  4. 2021BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation and Alignmentcited by 340
  5. 2020BasicVSR: The Search for Essential Components in Video Super-Resolution and Beyondcited by 303
  6. 2022VRT: A Video Restoration Transformercited by 219
  7. 2022Recurrent Video Restoration Transformer with Guided Deformable Attentioncited by 168
  8. 2021On Feature Decorrelation in Self-Supervised Learningcited by 160
  9. 2016Learning a No-Reference Quality Metric for Single-Image Super-Resolutioncited by 143
  10. 2017Fast and Accurate Image Super-Resolution with Deep Laplacian Pyramid Networkscited by 138
  11. 2021Investigating Tradeoffs in Real-World Video Super-Resolutioncited by 133
  12. 2020Unfolding the Alternating Optimization for Blind Super Resolutioncited by 129
  13. 2021Video Super-Resolution Transformercited by 120
  14. 2020Video Super-Resolution with Recurrent Structure-Detail Networkcited by 116
  15. 2015Reducing Overfitting in Deep Networks by Decorrelating Representationscited by 94
  16. 2022Rethinking Alignment in Video Super-Resolution Transformerscited by 71
  17. 2019Efficient Video Super-Resolution through Recurrent Latent Space Propagationcited by 69
  18. 2023Omni Aggregation Networks for Lightweight Image Super-Resolutioncited by 62
  19. 2020Deep Blind Video Super-resolutioncited by 44
  20. 2022Mitigating Artifacts in Real-World Video Super-Resolution Modelscited by 29
  21. 2023Expanding Synthetic Real-World Degradations for Blind Video Super Resolutioncited by 12
  22. 2023NegVSR: Augmenting Negatives for Generalized Noise Modeling in Real-World Video Super-Resolutioncited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 20; 2026: 13.

33 papers.

Sort:

By area:

  1. 2025SeedVR: Seeding Infinity in Diffusion Transformer Towards Generic Video Restorationcited by 62
  2. 2025STAR: Spatial-Temporal Augmentation with Text-to-Video Models for Real-World Video Super-Resolutioncited by 56
  3. 2025SeedVR2: One-Step Video Restoration via Diffusion Adversarial Post-Trainingcited by 45
  4. 2025FlashVSR: Towards Real-Time Diffusion-Based Streaming Video Super-Resolutioncited by 40
  5. 2025DOVE: Efficient One-Step Diffusion Model for Real-World Video Super-Resolutioncited by 30
  6. 2025UltraGen: High-Resolution Video Generation with Hierarchical Attentioncited by 17
  7. 2025InfVSR: Toward Consistency-Driven Streaming Generative Video Super-Resolutioncited by 8
  8. 2025Stream-DiffVSR: Low-Latency Streamable Video Super-Resolution via Auto-Regressive Diffusioncited by 7
  9. 2025Self-supervised ControlNet with Spatio-Temporal Mamba for Real-world Video Super-resolutioncited by 5
  10. 2025SimpleGVR: A Simple Baseline for Latent-Cascaded Video Super-Resolutioncited by 5
  11. 2025Blind Video Super-Resolution based on Implicit Kernelscited by 4
  12. 2025LiftVSR: Lifting Image Diffusion to Video Super-Resolution via Hybrid Temporal Modeling with Only 4×RTX 4090scited by 4
21 more
  1. 2026DUO-VSR: Dual-Stream Distillation for One-Step Video Super-Resolutioncited by 4
  2. 2025OS-DiffVSR: Towards One-step Latent Diffusion Model for High-detailed Real-world Video Super-Resolutioncited by 3
  3. 2025TurboVSR: Fantastic Video Upscalers and Where to Find Themcited by 2
  4. 2025Towards Redundancy Reduction in Diffusion Models for Efficient Video Super-Resolutioncited by 2
  5. 2026SwiftVR: Real-Time One-Step Generative Video Restorationcited by 2
  6. 2025ICME 2025 Grand Challenge on Video Super-Resolution for Video Conferencingcited by 1
  7. 2025Trajectory-aware Shifted State Space Models for Online Video Super-Resolutioncited by 1
  8. 2025CreativeVR: Diffusion-Prior-Guided Approach for Structure and Motion Restoration in Generative and Real Videoscited by 1
  9. 2026Taming Real-World Space-Time Video Super-Resolution with One-Step Diffusioncited by 1
  10. 2025Oneta: Multi-Style Image Enhancement Using Eigentransformation Functionscited by 0
  11. 2025STCDiT: Spatio-Temporally Consistent Diffusion Transformer for High-Quality Video Super-Resolutioncited by 0
  12. 2026Physics Encoded Spatial and Temporal Generative Adversarial Network for Tropical Cyclone Image Super-resolutioncited by 0
  13. 2026Compressed-Domain-Aware Online Video Super-Resolutioncited by 0
  14. 2026TextOVSR: Text-Guided Real-World Opera Video Super-Resolutioncited by 0
  15. 2026DTG-Restore: Training-Free Diffusion Refinement for Generative Video Super-Resolutioncited by 0
  16. 2026Binarized High-Efficiency RAW Video Restoration and Beyondcited by 0
  17. 2026ClearText-Video: A Large-Scale Text-Centric Video Dataset Bridging Video Restoration and Scene-Text Enhancementcited by 0
  18. 2026High Dynamic Range Video Reconstruction from Single-Exposure Raw Sequencescited by 0
  19. 2026LoCoVSR: Local Context Diffusion Posterior Sampling for Video Super-Resolutioncited by 0
  20. 2026ReCaVSR: One-Step Streaming Diffusion Video Super-Resolution with Recycled Latents and Learned Cache Routingcited by 0
  21. 2026RelayVSR: Large-Small Model Collaboration for Efficient Real-World Video Super-Resolutioncited by 0