Research map / Image generation / Image restoration
It Takes Two: A Duet of Periodicity and Directionality for Burst Flicker Removal
Lishen Qu, Shihao Zhou, Jie Liang, Hui Zeng, Lei Zhang, Jufeng Yang
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 PFM (phase-based fusion) in place of AST's CNN fusion (ablation) | Flickerformer / AST baseline · BurstDeflicker benchmark | PSNR (dB): 30.449 -> 30.728 (+0.279 dB) | Improved |
| Add AFFN (autocorrelation FFN) in place of FRFN (ablation) | Flickerformer / AST baseline · BurstDeflicker benchmark | PSNR (dB): 30.449 -> 30.831 (+0.382 dB) | Improved |
| Add WDAM (wavelet directional attention) in place of ASSA (ablation) | Flickerformer / AST baseline · BurstDeflicker benchmark | PSNR (dB): 30.449 -> 30.822 (+0.373 dB) | Improved |
| Combine PFM, AFFN and WDAM (ablation) | AST baseline · BurstDeflicker benchmark | PSNR (dB): 30.449 -> 31.226 | Improved |
| Use AFFN instead of FRFN (ablation) | Flickerformer · BurstDeflicker benchmark | PSNR (dB): 30.961 -> 31.226 (about +0.265 dB) | Improved |
| Use AFFN instead of GDFN (ablation) | Flickerformer · BurstDeflicker benchmark | PSNR (dB): 30.959 -> 31.226 | Improved |
| Use WDAM instead of conventional self-attention (Swin SA, Top-k SA, Condensed SA, ASSA) (ablation) | Flickerformer · BurstDeflicker benchmark | PSNR (dB): 30.997 -> 31.226 (+0.229 dB over best alternative) | Improved |
| Perform attention on the Haar low-frequency subband (WDAM) (ablation) | Flickerformer · BurstDeflicker benchmark | FLOPs (G): 139.42 -> 128.76 | Improved |
| Use Flickerformer instead of generic restoration networks | AST (second best) · BurstDeflicker benchmark | PSNR (dB): 30.646 -> 31.226 (+0.580 dB) | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Restormer — Restormer: Efficient Transformer for High-Resolution Image Restoration (CVPR 2022) · better on 2 of 2 shared results
- Burstormer — Burstormer: Burst Image Restoration and Enhancement Transformer (CVPR 2023) · 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
- BurstDeflicker: A Benchmark Dataset for Flicker Removal in Dynamic Scenes (NeurIPS 2025 · similarity 0.93)
- Bricker to BRACE: A Bracket Exposure RAW Dataset and Restoration Model for Flicker-Banding (NeurIPS 2026 · similarity 0.89)
- RIFLE: Removal of Image Flicker-Banding via Latent Diffusion Enhancement (NeurIPS 2026 · similarity 0.89)
- Interpretable Unsupervised Joint Denoising and Enhancement for Real-World low-light Scenarios (ICLR 2025 · similarity 0.83)
- Unveiling Details in the Dark: Simultaneous Brightening and Zooming for Low-Light Image Enhancement (AAAI 2024 · similarity 0.83)
- VQCNIR: Clearer Night Image Restoration with Vector-Quantized Codebook (AAAI 2024 · similarity 0.83)
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.
