Skip to content

Latest commit

 

History

History
53 lines (45 loc) · 11.1 KB

iccv21.md

File metadata and controls

53 lines (45 loc) · 11.1 KB

Paper with Code --- [ICCV2021]

Image Manipulation

  • Image Super-Resolution

    1. Fourier Space Losses for Efficient Perceptual Image Super-Resolution, Frequency loss, [Paper], [Code]
    2. Designing a Practical Degradation Model for Deep Blind Image Super-Resolution, Degradation Modeling, [Paper], [Code]
    3. Dynamic High-Pass Filtering and Multi-Spectral Attention for Image Super-Resolution, DCT + Model Design, [Paper], [Code]
    4. Context Reasoning Attention Network for Image Super-Resolution, Attention + Model Design, [Paper], [Code]
    5. Mutual Affine Network for Spatially Variant Kernel Estimation in Blind Image Super-Resolution, Spatially Variant Kernel Estimation+ SFT + Blind SR, [Paper], [Code]
    6. Benchmarking Ultra-High-Definition Image Super-resolution, UHD Datasets(4K/8K), [Paper], [Code]
    7. Attention-based Multi-Reference Learning for Image Super-Resolution, ReferenceSR, [Paper], [Code]
    8. Hierarchical Conditional Flow: A Unified Framework for Image Super-Resolution and Image Rescaling, Normalizing Flow + Image Rescaling + Reverse, [Paper], [Code]
    9. Learning A Single Network for Scale-Arbitrary Super-Resolution, Scale-Arbitrary, [Paper], [Code]
    10. Achieving on-Mobile Real-Time Super-Resolution with Neural Architecture and Pruning Search, Efficient SR, NAS, [Paper], [Code]
    11. Learning Frequency-aware Dynamic Network for Efficient Super-Resolution, Frequency, EfficientSR, [Paper], [Code]
    12. Super-Resolving Cross-Domain Face Miniatures by Peeking at One-Shot Exemplar, Face SR, [Paper], [Code]
    13. Lucas-Kanade Reloaded: End-to-End Super-Resolution from Raw Image Bursts, Raw Image SR, [Paper], [Code]
    14. Unsupervised Real-World Super-Resolution: A Domain Adaptation Perspective, Unpaired SR, [Paper], [Code]
    15. Dual-Camera Super-Resolution with Aligned Attention Modules, Dual-Camera SR, New Dataset, [Paper], [Code]

  • Image DeBlurring

    1. Single Image Defocus Deblurring Using Kernel-Sharing Parallel Atrous Convolutions, Defocus Deblurring, [Paper], [Code]
    2. Defocus Map Estimation and Deblurring from a Single Dual-Pixel Image, Defocus Map Estimation, Dual-Pixel Image, [Paper], [Code]
    3. Rethinking Coarse-to-Fine Approach in Single Image Deblurring, Coarse-to-Fine, Multi-scale, [Paper], [Code]
    4. Pyramid Architecture Search for Real-Time Image Deblurring, Pyramid Net, [Paper], [Code]
    5. Perceptual Variousness Motion Deblurring with Light Global Context Refinement, Perceptual Variousness Block, [Paper], [Code]

  • Image Denoising

    1. Rethinking Noise Synthesis and Modeling in Raw Denoising, Raw Denoising, [Paper], [Code]
    2. Hyperspectral Image Denoising with Realistic Data, Hyperspectral Image Denoising, BIT-Fuyin, [Paper], [Code]
    3. Cross-Patch Graph Convolutional Network for Image Denoising, Model Design, GCN, [Paper], [Code]
    4. End-to-End Unsupervised Document Image Blind Denoising, Document Image, Blind Denoising, [Paper], [Code]
    5. C2N: Practical Generative Noise Modeling for Real-World Denoising, Cycle GAN, Unpaired, [Paper], [Code]
    6. Rethinking Deep Image Prior for Denoising, DIP, [Paper], [Code]

Video Manipulation

  • Video SR

    1. Omniscient Video Super-Resolution, [Paper], [Code]
    2. COMISR: Compression-Informed Video Super-Resolution, Compressed Frame VSR, [Paper], [Code]
    3. Deep Blind Video Super-resolution, Blind VSR, [Paper], [Code]
    4. Real-world Video Super-resolution: A Benchmark Dataset and A Decomposition based Learning Scheme, New dataset, [Paper], [Code]
    5. Efficient Video Compression via Content-Adaptive Super-Resolution, Video Compression, VSR, [Paper], [Code]

  • Video Deblurring

    1. Multi-Scale Separable Network for Ultra-High-Definition Video Deblurring, Multi-Scale, [Paper], [Code]

  • Video Denosing

    1. Unsupervised Deep Video Denoising, not require explicit motion compensation, [Paper], [Code]
    2. Deep Reparametrization of Multi-Frame Super-Resolution and Denoising, Deep Reparametrization, Burst SR/Denoising, [Paper], [Code]
    3. Patch Craft: Video Denoising by Deep Modeling and Patch Matching, Patch Matching, Non-local, [Paper], [Code]