Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
A Multi-Mode Modulator for Multi-Domain Few-Shot Classification
Yanbin Liu, Juho Lee, Linchao Zhu, Ling Chen, Humphrey Shi, Yi Yang
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 domain-cooperative set to domain-specific set (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 71.9 -> 73.9 | Improved |
| add domain-specific set to domain-cooperative set (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 73.0 -> 73.9 | Improved |
| use cooperative instead of specific parameters on Omniglot (ablation) | tri-M · Omniglot | accuracy: 88.2 -> 90.5 | Improved |
| use cooperative instead of specific parameters on QuickDraw (ablation) | tri-M · QuickDraw | accuracy: 75.1 -> 76.2 | Improved |
| use 8 modes instead of 4 domain-specific and 4 cooperative modes (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 72.7 -> 73.9 | Improved |
| increase cooperative modes from 8 to 16 (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 73.9 -> 73.0 | Hurt |
| increase both mode counts from 8 to 16 (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 73.9 -> 72.2 | Hurt |
| add modulation on shallowest layer group (3 to 4 groups) (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 71.7 -> 73.9 | Improved |
| modulate 4 layer groups instead of 1 (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 68.3 -> 73.9 | Improved |
| modulate all layers instead of only second conv of each block (ablation) | tri-M · Meta-Dataset (13 datasets) | accuracy: 73.1 -> 73.9 | No clear effect |
| tri-M versus SimpleCNAPS on unseen domains | tri-M · Meta-Dataset out-of-domain | accuracy: 69.7 -> 72.9 | Improved |
Closest papers by meaning
- Selecting Relevant Features from a Multi-domain Representation for Few-shot Classification (ECCV 2020 · similarity 0.89)
- Discriminative Sample-Guided and Parameter-Efficient Feature Space Adaptation for Cross-Domain Few-Shot Learning (CVPR 2024 · similarity 0.88)
- Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation (ICLR 2020 · similarity 0.88)
- Cross-domain Few-shot Classification via Invariant-content Feature Reconstruction (IJCV 2026 · similarity 0.88)
- Switch to Generalize: Domain-Switch Learning for Cross-Domain Few-Shot Classification (ICLR 2022 · similarity 0.88)
- Mixup-Induced Domain Extrapolation for Domain Generalization (AAAI 2024 · similarity 0.86)
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.
