Research map / Learning theory and algorithms / Matrix and tensor methods
Online Convolutional Sparse Coding with Sample-Dependent Dictionary
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 sample-dependent filters formed from learned base filters | SCSC versus OCSC · Video | PSNR (dB): 33.17$\pm$0.01 -> 38.02$\pm$0.03 | Helped |
| Use sample-dependent filters formed from learned base filters | SCSC versus OCSC · Multispectral | PSNR (dB): 30.12$\pm$0.02 -> 31.71$\pm$0.01 | Helped |
| Use sample-dependent filters formed from learned base filters | SCSC versus CCSC · Light field | PSNR (dB): 13.70$\pm$0.09 -> 31.70$\pm$0.02 | Helped |
| Increase base-filter count from R=5 to R=10 (ablation) | SCSC · Video | PSNR (dB): 35.30$\pm$0.02 -> 38.02$\pm$0.03 | Helped |
| Increase base-filter count from R=5 to R=10 (ablation) | SCSC · Multispectral | PSNR (dB): 30.51$\pm$0.02 -> 31.71$\pm$0.01 | Helped |
| Increase base-filter count from R=5 to R=10 (ablation) | SCSC · Light field | PSNR (dB): 29.30$\pm$0.03 -> 31.70$\pm$0.02 | Helped |
| Use dense L2-constrained combination weights instead of L1-constrained weights (ablation) | SCSC · Fruit and City | PSNR | Helped |
| Make combination weights shared across samples instead of sample-dependent (ablation) | SCSC · Fruit and City | PSNR | Hurt |
| Post-process a shared OCSC dictionary into separable filters | SCSC comparison · Fruit and City | PSNR | Helped |
| Fine-tune shared OCSC filters for each test sample (ablation) | OCSC · Fruit and City | PSNR | Helped |
| Use SCSC rather than OCSC | Image denoising · Ten test images | PSNR (dB) | Helped |
| Use SCSC rather than OCSC | Image inpainting · Ten test images | PSNR (dB) | Mixed |
Closest papers by meaning
- Rethinking the CSC Model for Natural Images (NeurIPS 2019 · similarity 0.89)
- A Local Block Coordinate Descent Algorithm for the CSC Model (CVPR 2019 · similarity 0.88)
- DICOD: Distributed Convolutional Coordinate Descent for Convolutional Sparse Coding (ICML 2018 · similarity 0.87)
- A Provable Approach for Double-Sparse Coding (AAAI 2018 · similarity 0.86)
- SparseFormer: Sparse Visual Recognition via Limited Latent Tokens (ICLR 2024 · similarity 0.83)
- Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation (ICML 2025 · 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.