Research map / Learning theory and algorithms / Matrix and tensor methods
Solving Vision Problems via Filtering
Sean I. Young, Aous T. Naman, Bernd Girod, David Taubman
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 |
|---|---|---|---|
| replace conjugate-gradient solve with non-iterative filtering (geodesic) | geodesic regularizer · Depth SR 16x | running time: 1.60s -> 0.44s | Improved |
| replace conjugate-gradient solve with non-iterative filtering (bilateral) | bilateral regularizer · Depth SR 16x | running time: 8.23s -> 1.41s | Improved |
| replace Bilateral Solver with bilateral filtering | Bilateral Solver · Depth SR 16x | PSNR (dB): 35.15 -> 35.76 | Improved |
| replace SOR solver with geodesic filtering | EpicFlow variational optimizer · MPI-Sintel optical flow | running time: 1.19s -> 0.65s | Improved |
| replace SOR solver with geodesic filtering | EpicFlow variational optimizer · MPI-Sintel optical flow | average EPE: 0.784 -> 0.784 | No clear effect |
| replace iterative bilateral solver with bilateral filtering | iterative bilateral · MPI-Sintel optical flow | running time: 14.4s -> 3.32s | Improved |
| replace TV iterative solver with geodesic filtering | TV · Kodak deblurring | running time: 1.46s -> 0.07s | Improved |
| replace TV iterative solver with geodesic filtering | TV · Kodak deblurring | average PSNR: 31.77 -> 31.89 | No clear effect |
| use bilateral instead of geodesic filter | ours · Depth SR 16x | PSNR (dB): 35.32 -> 35.76; time 0.44s -> 1.41s | Mixed |
| use bilateral instead of geodesic filter | ours · Kodak deblurring | average PSNR: 31.89 -> 31.87 | No clear effect |
| use rectified generalized inverse instead of thresholded pseudo-inverse (ablation) | deblurring filter · Kodak deblurring | ringing artifacts | Improved |
Closest papers by meaning
- A Variational Perspective on Solving Inverse Problems with Diffusion Models (ICLR 2024 · similarity 0.85)
- Weakly Convex Regularisers for Inverse Problems: Convergence of Critical Points and Primal-Dual Optimisation (ICML 2024 · similarity 0.85)
- DEALing with Image Reconstruction: Deep Attentive Least Squares (ICML 2025 · similarity 0.85)
- Solving Video Inverse Problems Using Image Diffusion Models (ICLR 2025 · similarity 0.85)
- DiracDiffusion: Denoising and Incremental Reconstruction with Assured Data-Consistency (ICML 2024 · similarity 0.84)
- Learning Cocoercive Conservative Denoisers via Helmholtz Decomposition for Poisson Imaging Inverse Problems (NeurIPS 2025 · similarity 0.84)
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.
