AI Paper Atlas

Research map / Benchmarks / FMNIST

FMNIST leaderboard

28 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on FMNIST. Best accuracy: 99.21 (Lineage Detector, CVPR 2024).

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.

FMNIST · accuracy

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

#accuracy ↑MethodPaperTopic
199.21Lineage Detector
ResNet18
Neural Lineage
CVPR 2024 · Table 1
Open-set and few-shot detection
295.88CordsNet-R8 fine-tunedRecurrent neural network dynamical systems for biological vision
NeurIPS 2024 · Table 1
RNNs and normalization
395.4Dist-PU
6-layer MLP
Dist-PU: Positive-Unlabeled Learning From a Label Distribution Perspective
CVPR 2022 · Table: Comparative results
Semi-supervised learning
494.83Mapping Network (layer-wise, 4078 params)
CNN1
Mapping Networks
CVPR 2026 · Table 1
Variational and generative sampling
594.7VGG-11 trained with bias, test all
VGG-11
Bias Also Matters: Bias Attribution for Deep Neural Network Explanation
ICML 2019 · Table 1
RNNs and normalization
694.6gIInstance-wise Feature Grouping
NeurIPS 2020 · Table 2
Recognition and long-tail
794.34Spherized (last FC replaced)
VGG-11
Spherization Layer: Representation Using Only Angles
NeurIPS 2022 · Table 1
RNNs and normalization
893.6S2: train Blocks 2 and 4
Proposed CNN
Deep Network Approximation in Terms of Intrinsic Parameters
ICML 2022 · Test accuracy comparison table
Sample complexity
993.54LGL-INR
CNN5C
On the Learning Property of Logistic and Softmax Losses for Deep Neural Networks
AAAI 2020 · Predictive results table
Recognition and long-tail
1093.01SGHMC-SAAn Adaptive Empirical Bayesian Method for Sparse Deep Learning
NeurIPS 2019 · Table 2
Variational and generative sampling
18 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.