EMNIST: Extending MNIST to handwritten letters

By Gregory Cohen, Saeed Afshar, Jonathan C. Tapson, A. van Schaik. 2017

1,889 citations recorded by the source corpus.

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

In Computer Vision and Pattern Recognition

In Evaluation & Benchmarks

Citations in the corpus

Practical and Private (Deep) Learning without Sampling or Shuffling (2021, background): “, 1998), CIFAR10 (Krizhevsky, 2009), EMNIST (ByMerge split) (Cohen et al., 2017); and a next word prediction task on StackOverflow data set (Overflow, 2018).”
The Distributed Discrete Gaussian Mechanism for Federated Learning with Secure Aggregation (2021, methodology): “We evaluate on two realistic FL tasks: Federated EMNIST (Cohen et al., 2017; Caldas et al., 2018) and Stack 6The kinks on the low bit-width curves are due to the TensorFlow implementation of the discrete Gaussian sampler taking integer noise scales; to preserve privacy, noise scales are rounded up a”
Attack of the Tails: Yes, You Really Can Backdoor Federated Learning (2020, methodology): “Tasks We consider the following five tasks with various values of K (num. of clients) and m (num. of clients in each iteration): (Task 1) Image classification on CIFAR-10 [77] with VGG-9 [78] (K = 200,m = 10), (Task 2) Digit classification on EMNIST [79] with LeNet [80] (K = 3383,m = 30), (Task 3) I”
Adaptive Federated Optimization (2020, methodology): “We use federated versions of vision datasets EMNIST (Cohen et al., 2017) and CIFAR-100 (Krizhevsky and Hinton, 2009), and language modeling datasets Shakespeare (McMahan et al., 2017) and StackOverflow3.”
Can You Really Backdoor Federated Learning? (2019, methodology): “In the above backdoor attack framework, we conduct experiments on the EMNIST dataset (11; 9).”
Adaptive Personalized Federated Learning (2020, background): “, 2009), EMNIST (Cohen et al., 2017), and a synthetic dataset.”
Federated Visual Classification with Real-World Data Distribution (2020, methodology): “[2] use the Extended MNIST dataset [3] split”
A comprehensive, application-oriented study of catastrophic forgetting in DNNs (2019, background): “EMNIST (Cohen et al., 2017) is an extended version of MNIST with additional classes of hand-written letters.”
Addressing Loss of Plasticity and Catastrophic Forgetting in Continual Learning (2024, methodology): “In this section, we study the interplay of catastrophic forgetting and loss of plasticity using the Label-permuted EMNIST problem.”
Towards Probabilistic Verification of Machine Unlearning (2020, background): “Not Applicable digits form of EMNIST, which has 10 class labels, each with 280,000 samples [36].”
Regular Polytope Networks (2021, methodology): “conducted with the well-known MNIST, FashionMNIST [40], EMNIST [41], CIFAR-10, CIFAR-100 [42], and ImageNet”
On Anytime Learning at Macroscale (2021, result): “Finally, since findings are similar and because of the limited space, we report in Appendix A and B results on two other computer vision datasets, namely EMNIST [Cohen et al., 2017] and CIFAR 100 [Krizhevsky, 2009].”
Data-Free Knowledge Distillation for Heterogeneous Federated Learning (2021, methodology): “Effects of communication frequency: We explore different local updating steps T on the EMNIST, so that a higher T means longer communication delays before the global communication.”