Research map / Learning theory and algorithms / Matrix and tensor methods
HiNCoT: Hierarchical Nonlinear Continuous Transform-based Tensor Representation for Multi-Dimensional Data Recovery
Tao Yang, Weihao Wu, Tingzhu Huang
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 the low-rank core tensor as input to the deep representation module (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 43.448 -> 45.188 | Improved |
| add the deep representation module (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 41.480 -> 45.188 | Improved |
| add the HNC-transform module (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 44.403 -> 45.188 | Improved |
| use hierarchical nonlinear transform (l=3) instead of single-layer linear (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 42.972 -> 45.188 | Improved |
| add nonlinear activations to hierarchical transform (l=3) (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 43.059 -> 45.188 | Improved |
| increase depth from l=2 to l=3 (nonlinear) (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 43.838 -> 45.188 | Improved |
| increase depth from l=3 to l=4 (nonlinear) (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB): 45.188 -> 43.318 | Hurt |
| add low-rank core tensor, SR=0.02 (ablation) | HiNCoT · Flowers random missing SR=0.02 | PSNR (dB): 31.628 -> 32.338 | Improved |
| add deep representation module, SR=0.02 (ablation) | HiNCoT · Flowers random missing SR=0.02 | PSNR (dB): 29.097 -> 32.338 | Improved |
| add hierarchical nonlinear transform (l=3) vs single linear, SR=0.02 (ablation) | HiNCoT · Flowers random missing SR=0.02 | PSNR (dB): 31.248 -> 32.338 | Improved |
| increase tubal rank r from 1 to 30 (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB) | Mixed |
| increase core size b from 0.1 n3 to n3 (ablation) | HiNCoT · Flowers random missing SR=0.10 | PSNR (dB) | Mixed |
Closest papers by meaning
- DELTA: Deep Low-Rank Tensor Representation for Multi-Dimensional Data Recovery (TPAMI 2025 · similarity 0.91)
- Self-Attention Driven Tensor Representation for High-Order Data Recovery (CVPR 2026 · similarity 0.88)
- Low-Rank Tensor Function Representation for Multi-Dimensional Data Recovery (TPAMI 2023 · similarity 0.88)
- HLRTF: Hierarchical Low-Rank Tensor Factorization for Inverse Problems in Multi-Dimensional Imaging (CVPR 2022 · similarity 0.87)
- Deep Rank-One Tensor Functional Factorization for Multi-Dimensional Data Recovery (AAAI 2025 · similarity 0.87)
- Low-Rank Tensor Transitions (LoRT) for Transferable Tensor Regression (ICML 2025 · similarity 0.85)
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.
