Fashion-MNIST

Introduced by Xiao et al. in Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms

Fashion-MNIST is a dataset comprising of 28×28 grayscale images of 70,000 fashion products from 10 categories, with 7,000 images per category. The training set has 60,000 images and the test set has 10,000 images. Fashion-MNIST shares the same image size, data format and the structure of training and testing splits with the original MNIST.

Source: Generative Probabilistic Novelty Detection with Adversarial Autoencoders

Homepage

Benchmarks

Add a new result Link an existing benchmark

Task	Dataset Variant	Best Model
Image Classification	Fashion-MNIST	Fine-Tuning DARTS
Anomaly Detection	Fashion-MNIST	GAN based Anomaly Detection in Imbalance Problems
Image Clustering	Fashion-MNIST	AE+SNNL
Image Generation	Fashion-MNIST	PAE
Clustering Algorithms Evaluation	Fashion-MNIST	AE+GIT
Unsupervised Anomaly Detection with Specified Settings -- 10% anomaly	Fashion-MNIST	IF
Unsupervised Anomaly Detection with Specified Settings -- 1% anomaly	Fashion-MNIST	Isolation Forest
Unsupervised Anomaly Detection with Specified Settings -- 30% anomaly	Fashion-MNIST	IF
Unsupervised Anomaly Detection with Specified Settings -- 20% anomaly	Fashion-MNIST	IF
Unsupervised Anomaly Detection with Specified Settings -- 0.1% anomaly	Fashion-MNIST	Isolation Forest
Out-of-Distribution Detection	Fashion-MNIST	PAE
Domain Generalization	Rotated Fashion-MNIST	CSD
Multiview Clustering	Fashion-MNIST	MCoCo
Unsupervised Anomaly Detection	Fashion-MNIST	RSRAE
General Classification	Fashion-MNIST	CAE
Model Poisoning	Fashion-MNIST	ADA
Outlier Detection	Fashion-MNIST	PAE