Research map / Label-efficient and robust learning / Recognition and long-tail
Few-shot Learning for Feature Selection with Hilbert-Schmidt Independence Criterion
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 |
|---|---|---|---|
| add latent task representation z to kernels (vs w/o Latent) (ablation) | Ours · Avg over datasets | test accuracy [%]: 70.0 -> 70.7 | Helped |
| add latent task representation z to kernels (vs w/o Latent) (ablation) | Ours · IoT | test accuracy [%]: 74.0 -> 72.6 | Hurt |
| use task-specific feature-invariant kernel (w/o Feature) (ablation) | Ours · Avg over datasets | test accuracy [%]: 69.3 -> 70.7 | Helped |
| use feature-specific kernel parameters (vs w/o LFeature) (ablation) | Ours · Avg over datasets | test accuracy [%]: 68.8 -> 70.7 | Helped |
| use S-dependent kernel (vs w/o S-kernel) (ablation) | Ours · Avg over datasets | test accuracy [%]: 68.9 -> 70.7 | Helped |
| learn initial coefficient parameters (vs w/o Initial) (ablation) | Ours · Avg over datasets | test accuracy [%]: 68.8 -> 70.7 | Helped |
| use proposed kernel form instead of deep kernel (w/ Deep) (ablation) | Ours · Avg over datasets | test accuracy [%]: 68.4 -> 70.7 | Helped |
| perform support set adaptation (vs w/o S-adapt) (ablation) | Ours · Avg over datasets | test accuracy [%]: 69.3 -> 70.7 | Helped |
| support set adaptation (vs w/o S-adapt) (ablation) | Ours · IoT | test accuracy [%]: 73.4 -> 72.6 | Hurt |
| support set adaptation (vs w/o S-kernel) on Mnistr (ablation) | Ours · Mnistr | test accuracy [%]: 70.6 -> 71.4 | Helped |
| support set adaptation (vs w/o S-kernel) on Isolet (ablation) | Ours · Isolet | test accuracy [%]: 65.6 -> 68.0 | Helped |
| add latent task representation (vs w/o Latent) (ablation) | Ours · Isolet | test accuracy [%]: 65.7 -> 68.0 | Helped |
Closest papers by meaning
- Balancing Feature Similarity and Label Variability for Optimal Size-Aware One-shot Subset Selection (ICML 2024 · similarity 0.85)
- MOKD: Cross-domain Finetuning for Few-shot Classification via Maximizing Optimized Kernel Dependence (ICML 2024 · similarity 0.84)
- SAND: One-Shot Feature Selection with Additive Noise Distortion (ICML 2025 · similarity 0.84)
- Rethinking Few-Shot Adaptation of Vision-Language Models in Two Stages (CVPR 2025 · similarity 0.84)
- Making Large Vision Language Models to Be Good Few-Shot Learners (AAAI 2025 · similarity 0.83)
- Boosting Few-Shot Learning via Attentive Feature Regularization (AAAI 2024 · 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.