Research map / Detection and segmentation / CNN features
R²MRF: Defocus Blur Detection via Recurrently Refining Multi-Scale Residual Features
Chang Tang, Xinwang Liu, Xinzhong Zhu, En Zhu, Kun Sun, Pichao Wang, Lizhe Wang, Albert Zomaya
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 residual refinement modules (residual learning) to the recurrent branches (ablation) | R2MRF · DUT | F-measure: 0.918 -> 0.950 | Improved |
| add residual refinement modules (ablation) | R2MRF · Shi et al. dataset | F-measure: 0.905 -> 0.927 | Improved |
| increase recurrent steps from 1 to 6 (ablation) | R2MRF · DUT | F-measure: 0.921 -> 0.950 | Improved |
| increase recurrent steps from 6 to 8 (ablation) | R2MRF · DUT | F-measure: 0.950 -> 0.952 | No clear effect |
| fuse side outputs of all branches instead of using a single branch output (ablation) | R2MRF · DUT | F-measure: 0.936 -> 0.950 | Improved |
| use ResNeXt backbone instead of VGG-16 (ablation) | R2MRF · DUT | F-measure: 0.932 -> 0.950 | Improved |
| use ResNeXt backbone instead of DenseNet-161 (ablation) | R2MRF · DUT | F-measure: 0.934 -> 0.950 | Improved |
| R2MRF vs BTBNet on Shi et al. dataset MAE | defocus blur detection · Shi et al. dataset | MAE: 0.105 -> 0.119 | Mixed |
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.
- DeFusionNET — DeFusionNET: Defocus Blur Detection via Recurrently Fusing and Refining Multi-Scale Deep Features (CVPR 2019) · better on 5 of 6 shared results
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
- DeFusionNET: Defocus Blur Detection via Recurrently Fusing and Refining Multi-Scale Deep Features (CVPR 2019 · similarity 0.96)
- DeFusionNET: Defocus Blur Detection via Recurrently Fusing and Refining Discriminative Multi-Scale Deep Features (TPAMI 2020 · similarity 0.95)
- Residual Diffusion Deblurring Model for Single Image Defocus Deblurring (AAAI 2025 · similarity 0.89)
- Multi-Focus Image Fusion via Explicit Defocus Blur Modelling (AAAI 2025 · similarity 0.88)
- Prior and Prediction Inverse Kernel Transformer for Single Image Defocus Deblurring (AAAI 2024 · similarity 0.88)
- Quad-Pixel Image Defocus Deblurring: A New Benchmark and Model (CVPR 2025 · similarity 0.87)
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.
