Research map / Video / Video segmentation and tracking
Spatio-Temporal Deformable Attention Network for Video Deblurring
Huicong Zhang, Haozhe Xie, Hongxun Yao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add MSA layer (with MMA absent) (ablation) | STDANet · GoPro | PSNR: 30.12 -> 31.15 | Improved |
| add MMA layer (with MSA absent) (ablation) | STDANet · GoPro | PSNR: 30.12 -> 31.18 | Improved |
| add both MMA and MSA layers instead of concatenation (ablation) | STDANet · GoPro | PSNR: 30.12 -> 32.29 | Improved |
| add both MMA and MSA layers (ablation) | STDANet · GoPro | SSIM: 0.8950 -> 0.9313 | Improved |
| increase sampling points from 1 to 12 (ablation) | STDANet · GoPro | PSNR: 31.64 -> 32.29 | Improved |
| increase sampling points from 12 to 16 (ablation) | STDANet · GoPro | PSNR: 32.29 -> 32.32 | No clear effect |
| increase attention heads from 1 to 4 (ablation) | STDANet · GoPro | PSNR: 32.13 -> 32.29 | Improved |
| increase attention heads from 4 to 8 (ablation) | STDANet · GoPro | GMACs: 1677 -> 1849 | Hurt |
| add lightweight motion estimator (ablation) | STDANet · GoPro | PSNR: 31.58 -> 32.29 | Improved |
| replace motion estimator with PWC-Net (ablation) | STDANet · GoPro | PSNR: 32.29 -> 32.36 | Improved |
| cascade STDANet into STDANet-Stack | STDANet · DVD | PSNR: 32.63 -> 33.05 | Improved |
| cascade STDANet into STDANet-Stack | STDANet · BSD (2ms-16ms) | PSNR: 33.13 -> 33.27 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ARVo — ARVo: Learning All-Range Volumetric Correspondence for Video Deblurring (CVPR 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Spatio-Temporal Filter Adaptive Network for Video Deblurring (ICCV 2019 · similarity 0.90)
- Real-World Video Deblurring: A Benchmark Dataset and an Efficient Recurrent Neural Network (IJCV 2022 · similarity 0.90)
- Gated Spatio-Temporal Attention-Guided Video Deblurring (CVPR 2021 · similarity 0.89)
- Blur-aware Spatio-temporal Sparse Transformer for Video Deblurring (CVPR 2024 · similarity 0.88)
- Motion Deblurring via Spatial-Temporal Collaboration of Frames and Events (AAAI 2024 · similarity 0.87)
- Learning Scale-Aware Spatio-temporal Implicit Representation for Event-based Motion Deblurring (ICML 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
