Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
DeepEMD: Few-Shot Image Classification With Differentiable Earth Mover's Distance and Structured Classifiers
Chi Zhang, Yujun Cai, Guosheng Lin, Chunhua Shen
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 |
|---|---|---|---|
| Use local feature representations with EMD matching | few-shot image classification · miniImageNet 1-shot 5-way | accuracy: 63.08 -> 65.91 | Improved |
| Use local feature representations with EMD matching | few-shot image classification · miniImageNet 1-shot 5-way | accuracy: 62.52 -> 65.91 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · miniImageNet 5-shot 5-way | accuracy: 80.51 -> 82.41 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · tieredImageNet 1-shot 5-way | accuracy: 68.50 -> 71.16 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · tieredImageNet 5-shot 5-way | accuracy: 84.28 -> 86.03 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · FC100 1-shot 5-way | accuracy: 45.10 -> 46.47 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · FC100 5-shot 5-way | accuracy: 57.6 -> 63.22 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · CUB 1-shot 5-way | accuracy: 71.87 -> 75.65 | Improved |
| Use DeepEMD instead of the strongest listed baseline | few-shot image classification · CUB 5-shot 5-way | accuracy: 85.08 -> 88.69 | Improved |
| Add cross-reference weighting to EMD node weights (ablation) | DeepEMD · miniImageNet 1-shot 5-way | accuracy: 56.95 -> 61.13 | Improved |
| Replace cross-reference weights with K-means weights (ablation) | DeepEMD · miniImageNet 1-shot 5-way | accuracy: 61.13 -> 56.25 | Hurt |
| Replace cross-reference weights with mean-shift weights (ablation) | DeepEMD · miniImageNet 1-shot 5-way | accuracy: 61.13 -> 53.56 | Hurt |
Later papers that compared with this method
25 later accepted papers report DeepEMD as a baseline in their results tables; 20 of them report a better number on most of the results they share with it.
- Contextualizing Meta-Learning via Learning to Decompose (TPAMI 2023 · better on 6 of 6 shared results)
- Frequency Guidance Matters in Few-Shot Learning (ICCV 2023 · better on 4 of 4 shared results)
- Learning Instance and Task-Aware Dynamic Kernels for Few-Shot Learning (ECCV 2022 · better on 6 of 6 shared results)
- Adaptive Poincaré Point to Set Distance for Few-Shot Classification (AAAI 2022 · better on 5 of 5 shared results)
- PatchMix Augmentation to Identify Causal Features in Few-Shot Learning (TPAMI 2022 · better on 4 of 4 shared results)
- Learngene: From Open-World to Your Learning Task (AAAI 2022 · better on 3 of 3 shared results)
- DMN4: Few-Shot Learning via Discriminative Mutual Nearest Neighbor Neural Network (AAAI 2022 · better on 2 of 2 shared results)
- How to Train Your MAML to Excel in Few-Shot Classification (ICLR 2022 · better on 1 of 2 shared results)
17 more not listed.
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
- DeepEMD: Differentiable Earth Mover's Distance for Few-Shot Learning (TPAMI 2022 · similarity 0.99)
- EMD Metric Learning (AAAI 2018 · similarity 0.87)
- Revisiting Local Descriptor Based Image-To-Class Measure for Few-Shot Learning (CVPR 2019 · similarity 0.87)
- Joint Distribution Matters: Deep Brownian Distance Covariance for Few-Shot Classification (CVPR 2022 · similarity 0.87)
- Variational Patch Gating for Training-Free Few-Shot Classification (ECCV 2026 · similarity 0.86)
- UNEM: UNrolled Generalized EM for Transductive Few-Shot Learning (CVPR 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.
