AI Paper Atlas

Research map / Benchmarks / Omniglot 5-way 1-shot

Omniglot 5-way 1-shot leaderboard

36 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on Omniglot 5-way 1-shot. Best accuracy: 99.97 (MC2 ensemble, NeurIPS 2019).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

Omniglot 5-way 1-shot · accuracy

36 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
199.97MC2 ensemble
MAML
Meta-Curvature
NeurIPS 2019 · Few-shot classification result
Zero-/few-shot and meta-learning
299.9MetaKernel
shallow CNN
MetaKernel: Learning Variational Random Features With Limited Labels
TPAMI 2022 · Table 1
Uncertainty and Gaussian processes
399.8MetaVRF bi-LSTMLearning to Learn Kernels with Variational Random Features
ICML 2020 · Table 2
Variational and generative sampling
499.72Ours
4-conv backbone
Few-Shot Learning With Global Class Representations
ICCV 2019 · Table 1
Zero-/few-shot and meta-learning
599.7VersaMeta-Learning Probabilistic Inference for Prediction
ICLR 2019 · Accuracy results table
Molecules and proteins
699.67Variational Few-Shot (ours)Variational Few-Shot Learning
ICCV 2019 · Table 1
Uncertainty and Gaussian processes
799.62MAML-stop
MAML
Learning To Stop While Learning To Predict
ICML 2020 · Table: Few-shot classification
RNNs and normalization
899.6Relation NetworkLearning to Compare: Relation Network for Few-Shot Learning
CVPR 2018 · Table 1
Zero-/few-shot and meta-learning
999.5MT-netGradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
ICML 2018 · Table 3
Optimizers
1099.47MAML++
MAML
How to train your MAML
ICLR 2019 · Table: MAML++ Omniglot 5-way R
RNNs and normalization
26 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.