Research map / Label-efficient and robust learning / Recognition and long-tail
A Unifying Framework in Vector-valued Reproducing Kernel Hilbert Spaces for Manifold Regularization and Co-Regularized Multi-view Learning
Hà Quang Minh, Loris Bazzani, Vittorio Murino
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 between-view regularization (γB=1e-6) (ablation) | MVL-LS · Caltech-101 (lc=1, uc=5) | accuracy: 30.59% -> 32.44% | Improved |
| add within-view regularization (γW=1e-6) (ablation) | MVL-LS · Caltech-101 (lc=1, uc=5) | accuracy: 30.59% -> 31.81% | Improved |
| add both between-view and within-view regularization (ablation) | MVL-LS · Caltech-101 (lc=1, uc=5) | accuracy: 30.59% -> 32.94% | Improved |
| add both regularizers (ablation) | MVL-LS · Caltech-101 (lc=uc=5) | accuracy: 63.68% -> 64.2% | Improved |
| combine 10 views instead of best single view PHOW gray L2 | MVL-LS · Caltech-101 (lc=1) | accuracy: 25.64% -> 30.41% | Improved |
| combine 10 views instead of best single view PHOW gray L2 | MVL-LS · Caltech-101 (lc=5) | accuracy: 56.75% -> 61.46% | Improved |
| combine PHOW and attributes views instead of PHOW only | MVL-LS · Caltech-UCSD Birds-200-2011 (lc=15) | accuracy: 9.92% -> 46.74% | Improved |
| optimize combination weights instead of uniform weights | MVL-LS-optC (25 it.) · Caltech-101 (10 kernels, lc=10) | accuracy: 68.1% -> 70.6% | Improved |
| optimize combination weights instead of uniform weights | MVL-LS-optC (25 it.) · Flower (lc=1) | accuracy: 39.41% -> 43.14% | Improved |
| optimize combination weights instead of uniform weights | MVL-LS-optC (25 it.) · Flower (lc=20) | accuracy: 81.76% -> 81.47% | No clear effect |
| increase optimization iterations from 25 to 50 | MVL-LS-optC · Caltech-101 (top 4 kernels, lc=5) | accuracy: 64.5% -> 64.7% | No clear effect |
| add unlabeled data (uc=5) in semi-supervised setup | MVL-LS · Flower (lc=1) | accuracy: 39.41% -> 41.86% | Improved |
Closest papers by meaning
- Localisation of Regularised and Multiview Support Vector Machine Learning (JMLR 2024 · similarity 0.89)
- Generalization Properties of hyper-RKHS and its Applications (JMLR 2021 · similarity 0.86)
- When VLAD Met Hilbert (CVPR 2016 · similarity 0.86)
- Multi-View Randomized Kernel Classification via Nonconvex Optimization (AAAI 2024 · similarity 0.86)
- Approximate Log-Hilbert-Schmidt Distances Between Covariance Operators for Image Classification (CVPR 2016 · similarity 0.86)
- Multi-View Oriented GPLVM: Expressiveness and Efficiency (NeurIPS 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.
