Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Generating Representative Samples for Few-Shot Classification
Jingyi Xu, Hieu Le
Check 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 CVAE-generated semantic-conditioned features | Meta-Baseline · miniImageNet 1-shot | accuracy (%): 63.17 ± 0.23 -> 69.96 ± 0.21 | Improved |
| Train the CVAE on representative samples instead of all base features (ablation) | Meta-Baseline + SVAE · miniImageNet 1-shot | accuracy (%): 69.96 ± 0.21 -> 72.79 ± 0.19 | Improved |
| Add CVAE-generated semantic-conditioned features | ProtoNet · miniImageNet 1-shot | accuracy (%): 62.39 -> 73.01 ± 0.24 | Improved |
| Train the CVAE on representative samples instead of all base features (ablation) | ProtoNet + SVAE · miniImageNet 1-shot | accuracy (%): 73.01 ± 0.24 -> 74.84 ± 0.23 | Improved |
| Add CVAE-generated semantic-conditioned features | E3BM · miniImageNet 1-shot | accuracy (%): 64.09 ± 0.37 -> 73.07 ± 0.39 | Improved |
| Train the CVAE on representative samples instead of all base features (ablation) | E3BM + SVAE · miniImageNet 1-shot | accuracy (%): 73.07 ± 0.39 -> 73.35 ± 0.37 | Improved |
| Add CVAE-generated semantic-conditioned features | Meta-Baseline · tieredImageNet 1-shot | accuracy (%): 68.62 ± 0.27 -> 73.05 ± 0.24 | Improved |
| Train the CVAE on representative samples instead of all base features (ablation) | Meta-Baseline + SVAE · tieredImageNet 1-shot | accuracy (%): 73.05 ± 0.24 -> 73.90 ± 0.24 | Improved |
| Add CVAE-generated semantic-conditioned features | ProtoNet · tieredImageNet 1-shot | accuracy (%): 68.23 -> 76.36 ± 0.65 | Improved |
| Train the CVAE on representative samples instead of all base features (ablation) | ProtoNet + SVAE · tieredImageNet 1-shot | accuracy (%): 76.36 ± 0.65 -> 76.98 ± 0.65 | Improved |
| Add CVAE-generated semantic-conditioned features | E3BM · tieredImageNet 1-shot | accuracy (%): 71.34 ± 0.41 -> 79.85 ± 0.43 | Improved |
| Train the CVAE on representative samples instead of all base features (ablation) | E3BM + SVAE · tieredImageNet 1-shot | accuracy (%): 79.85 ± 0.43 -> 80.46 ± 0.43 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Meta-Baseline — Meta-Baseline: Exploring Simple Meta-Learning for Few-Shot Learning (ICCV 2021) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Zero-Shot Learning via Class-Conditioned Deep Generative Models (AAAI 2018 · similarity 0.90)
- Few-Shot Learning With Embedded Class Models and Shot-Free Meta Training (ICCV 2019 · similarity 0.88)
- Synthesized Feature Based Few-Shot Class-Incremental Learning on a Mixture of Subspaces (ICCV 2021 · similarity 0.88)
- GenZSL: Generative Zero-Shot Learning Via Inductive Variational Autoencoder (ICML 2025 · similarity 0.88)
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.87)
- Envisioning Class Entity Reasoning by Large Language Models for Few-shot Learning (AAAI 2025 · similarity 0.87)
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.
