Research map / 3D vision / Depth and stereo
Dense Deep Unfolding Network With 3D-CNN Prior for Snapshot Compressive Imaging
Zhuoyuan Wu, Jian Zhang, Chong Mou
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 |
|---|---|---|---|
| use 3D-CNN instead of 2D-CNN in the prior (ablation) | plain UNet · Video SCI simulation (average) | PSNR: 32.77 -> 34.49 | Improved |
| add first DFM branch (DFM1) (ablation) | 3DCN · Video SCI simulation (average) | PSNR: 34.49 -> 34.54 | No clear effect |
| add second DFM branch (DFM2) (ablation) | 3DCN + DFM1 · Video SCI simulation (average) | PSNR: 34.54 -> 34.80 | Improved |
| add third DFM branch (DFM3) (ablation) | 3DCN + DFM1 + DFM2 · Video SCI simulation (average) | PSNR: 34.80 -> 34.94 | Improved |
| add all dense feature map branches (ablation) | 3DCN · Video SCI simulation (average) | PSNR: 34.49 -> 34.94 | Improved |
| add DFMA to all DFM branches (ablation) | 3DCN + DFM1-3 · Video SCI simulation (average) | PSNR: 34.94 -> 35.26 | Improved |
| add DFMA to DFM1 (ablation) | 3DCN + DFM1 · Video SCI simulation (average) | PSNR: 34.54 -> 34.81 | Improved |
| add DFMA to DFM1+DFM2 (ablation) | 3DCN + DFM1 + DFM2 · Video SCI simulation (average) | PSNR: 34.80 -> 35.03 | Improved |
| increase phase number to K=4 versus baseline RevSCI (ablation) | Dense DUN · Video SCI simulation (average) | PSNR: +0.26dB over RevSCI | Improved |
| use more phases (K up to 10) (ablation) | Dense DUN · Video SCI simulation (average) | PSNR | Improved |
| compare against best prior method RevSCI | Dense DUN · Video SCI simulation (average) | PSNR: 33.92 -> 35.26 | Improved |
Closest papers by meaning
- Using Powerful Prior Knowledge of Diffusion Model in Deep Unfolding Networks for Image Compressive Sensing (CVPR 2025 · similarity 0.89)
- Untrained Neural Nets for Snapshot Compressive Imaging: Theory and Algorithms (NeurIPS 2024 · similarity 0.89)
- Dual Prior Unfolding for Snapshot Compressive Imaging (CVPR 2024 · similarity 0.89)
- CPP-Net: Embracing Multi-Scale Feature Fusion into Deep Unfolding CP-PPA Network for Compressive Sensing (CVPR 2024 · similarity 0.88)
- HUNet: Homotopy Unfolding Network for Image Compressive Sensing (CVPR 2025 · similarity 0.87)
- UFC-Net: Unrolling Fixed-point Continuous Network for Deep Compressive Sensing (CVPR 2024 · 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.
