Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
By Han Xiao, Kashif Rasul, Roland Vollgraf. 2017
10,881 citations recorded by the source corpus.
We present Fashion-MNIST, a new dataset comprising of 28x28 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 is intended to serve as a direct drop-in replacement for the original MNIST dataset for benchmarking machine learning algorithms, as it shares the same image size, data format and the structure of training and testing splits. The dataset is freely available at this https URL
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Evaluation & Benchmarks
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Yoshua Bengio
- Sergey Levine
- Mohit Bansal
- Quoc V. Le
- Percy Liang
- Ruslan Salakhutdinov
- Jianfeng Gao
- Dacheng Tao
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Trevor Darrell
- Ming-Hsuan Yang
- Xiaogang Wang
- Jitendra Malik
- Yi Yang
- Bernt Schiele
In Evaluation & Benchmarks
- Jitendra Malik
- Frank Hutter
- Yi Yang
- Yoshua Bengio
- Kyunghyun Cho
- Timothée Lesort
- Ali Borji
- Kate Saenko
Citations in the corpus
- Measuring the Effects of Data Parallelism on Neural Network Training (2018, background): “10 Keskar et al. (2017) trained several neural network architectures on MNIST and CIFAR-10, each with two batch sizes, using the Adam optimizer and without changing the learning rate between batch sizes.”
- Toward Understanding Catastrophic Forgetting in Continual Learning (2019, methodology): “Similar constructions can also be applied to the not-MNIST set [7], the fashion MNIST set [48], or the CIFAR set [23] such as in the split not-MNIST [29] and split CIFAR benchmarks [51].”
- DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning (2021, positive): “In the rest of this section, we present experiments on the following real MTL datasets: MovieLens, Multi-MNIST, Multi-Fashion MNIST, and on a real-world, large-scale recommender system.”
- On Empirical Comparisons of Optimizers for Deep Learning (2019, methodology): “Simple CNN on Fashion MNIST used ( K, N ) = (100 , 500) ; ResNet-32 on CIFAR-100 used ( K, N ) = (100 , 500) ; ResNet-50 on ImageNet used ( K, N ) = (50 , 250) ; Transformer on LM1B used ( K, N ) = (50 , 100) ; VGG on CIFAR-10 with our code used ( K, N ) = (50 , 250) for tuning the learning rate sch”
- Stochastic Normalizing Flows (2020, background): “MNIST and Fashion-MNIST VAE in Table 1 • The latent space dimension was set to 50.”
- Energy-Inspired Models: Learning with Sampler-Induced Distributions (2019, methodology): “We evaluated the proposed models on a set of synthetic datasets, binarized MNIST [42] and Fashion MNIST [67], and continuous MINST, Fashion MNIST, and CelebA [44].”
- Diagnosing and Enhancing VAE Models (2019, methodology): “Note that the relatively poor performance of the WAE-MMD on MNIST and Fashion MNIST data can be attributed to the sensitivity of this approach to the value of κ , which for consistency with other models was fixed at κ = 64 for all experiments.”
- Sharpness-Aware Minimization for Efficiently Improving Generalization (2020, methodology): “, 2011) and Fashion-MNIST datasets (Xiao et al., 2017).”
- Negative Momentum for Improved Game Dynamics (2018, methodology): “6] In our third set of experiments, we use negative momentum in a GAN setup on CIFAR-10 (Krizhevsky and Hinton, 2009) and Fashion-MNIST (Xiao et al., 2017) with saturating loss and alternating steps.”
- UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction (2018, methodology): “F-MNIST [29] or Fashion MNIST is a dataset of 28x28 pixel grayscale images of fashion items (clothing, footwear and bags).”
- Assessing Generative Models via Precision and Recall (2018, methodology): “We consider three data sets commonly used in the GAN literature: MNIST [15], Fashion-MNIST [25], and CIFAR-10 [13].”
- Random Erasing Data Augmentation (2017, methodology): “For image classification, we evaluate on three image classification datasets, including two well-known datasets, CIFAR-10 and CIFAR-100 [13], and a new dataset Fashion-MNIST [28].”
- Differentially Private Learning Needs Better Features (or Much More Data) (2020, methodology): “, 2010), Fashion-MNIST (Xiao et al., 2017) and CIFAR-10 (Krizhevsky, 2009).”
- Tempered Sigmoid Activations for Deep Learning with Differential Privacy (2020, methodology): “Experimental Setup We use three common benchmarks for differentially private ML: MNIST (LeCun, Cortes, and Burges 1998), FashionMNIST (Xiao, Rasul, and Vollgraf 2017), and CIFAR10 (Krizhevsky, Hinton et al. 2009).”
- Flower: A Friendly Federated Learning Research Framework (2020, neutral): “There is active work on privacy and robustness improvements for FL: A targeted model poisoning attack using Fashion-MNIST (Xiao et al., 2017) (along with possible mitigation strategies) was demonstrated by Bhagoji et al. (2018).”