Research map / Learning theory and algorithms / Clustering
Sparse Subspace Clustering by Learning Approximation ℓ0 Codes
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 |
|---|---|---|---|
| approximate the OMP l0 code with a trained single-hidden-layer network (SSC-LOMP vs SSC-OMP) | SSC-OMP · MNIST (600 points/class) | ACC (%): 90.56±5.48 -> 94.32±0.72 | Helped |
| SSC-LOMP vs SSC-OMP on Extended YaleB | SSC-OMP · Extended YaleB (38 classes) | ACC (%): 77.60±1.95 -> 78.01±1.43 | Helped |
| SSC-LOMP vs SSC-OMP on 2-class Extended YaleB | SSC-OMP · Extended YaleB (2 classes) | ACC (%): 99.21±1.22 -> 99.09±1.28 | No clear effect |
| SSC-LOMP vs SSC-ISTA (l0 net vs l1 solver) | SSC-ISTA · Extended YaleB (38 classes) | ACC (%): 67.20±1.30 -> 78.01±1.43 | Helped |
| use shrinkage output activation instead of linear (ablation) | SSC-LOMP · MNIST (1000 points) | ACC (%): 77.37±9.80 -> 89.67±3.67 | Helped |
| use shrinkage output activation instead of ReLU (ablation) | SSC-LOMP · MNIST (1000 points) | ACC (%): 86.64±6.00 -> 89.67±3.67 | Helped |
| use shrinkage output activation instead of linear (ablation) | SSC-LOMP · Extended YaleB (10 classes) | ACC (%): 47.48±4.19 -> 90.47±4.43 | Helped |
| use shrinkage output activation instead of ReLU (ablation) | SSC-LOMP · Extended YaleB (10 classes) | ACC (%): 69.51±7.47 -> 90.47±4.43 | Helped |
| replace OMP at test time with SHNN forward pass | SSC-OMP · MNIST (600 points/class) | time (s): 35.9 -> 15.5 | Helped |
Closest papers by meaning
- Scalable Sparse Subspace Clustering by Orthogonal Matching Pursuit (CVPR 2016 · similarity 0.93)
- Stochastic Sparse Subspace Clustering (CVPR 2020 · similarity 0.90)
- Hyperplane Clustering via Dual Principal Component Pursuit (ICML 2017 · similarity 0.89)
- Neural Collaborative Subspace Clustering (ICML 2019 · similarity 0.89)
- Selective Sampling-based Scalable Sparse Subspace Clustering (NeurIPS 2019 · similarity 0.89)
- Exploring a Principled Framework for Deep Subspace Clustering (ICLR 2025 · similarity 0.88)
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.