NIST Special Database 19 Handprinted Forms and Characters Database

By Patrick Grother. 1995

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

EMNIST: an extension of MNIST to handwritten letters (2017, methodology): “The MNIST database is a subset of a much larger dataset known as the NIST Special Database 19 [11].”
Invariant Risk Minimization (2019, background): “For instance, whereas the original NIST handwritten data was collected from different writers under different conditions [19], the popular MNIST training and testing sets [8] were carefully shuffled to represent similar mixes of writers.”
Neurogenesis deep learning: Extending deep networks to accommodate new classes (2016, methodology): “We evaluated NDL on two datasets, the MNIST [9] and NIST SD 19 [10] datasets.”
Learning from one example through shared densities on transforms (2000, methodology): “We performed experiments on NIST Special Database 19 [7].”
Supervised learning of fuzzy ARTMAP neural networks through particle swarm optimization (2007, methodology): “Figure 20: Fuzzy ARTMAP generalization error versus training subset size on hsf-4 for the NIST SD19 data.”
Shape representation and classification using the Poisson equation (2004, methodology): “We first applied this procedure to numerals from the NIST special database 19 [12] (downsampled to fit in a 64×64 pixel box, centered about their centroid and brought to a uniform scale).”
Multi-column deep neural networks for image classification (2012, methodology): “Here we train such nets by simple online back-propagation, setting new, greatly improved records on MNIST [19], Latin letters [13], Chinese characters [22], traffic signs [33], NORB (jittered, cluttered) [20] and CIFAR10 [17] benchmarks.”
EMNIST: Extending MNIST to handwritten letters (2017, methodology): “The NIST Special Database 19 [3] contains handwritten digits and characters collected from over 500 writers.”
Transfer learning for Latin and Chinese characters with Deep Neural Networks (2012, methodology): “We experiment with Latin characters [4] and Chinese characters [5].”
Towards Probabilistic Verification of Machine Unlearning (2020, methodology): “The dataset is composed of handwritten character digits derived from the NIST Special Database 19 [35].”
Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarks (2021, methodology): “The dataset was constructed from Handwriting Sample Forms distributed to Census Bureau employees and high school students; the ground-truth labels were determined by matching digits to the instructions of the task to copy a particular set of digits [11].”
Estimating accurate multi-class probabilities with support vector machines (2005, background): “So, given a training set of instance-label pairs { }( , ) : 1, ,k kx y k n= … , where d kx ∈ and { }1, 1ky ∈ − , the unthresholded output of an SVM is 1 ( ) ( , ) n k k k k f x y K x x bα = = +∑ , ( 1 ) where the samples with non-zero Lagrange multiplier αk are called support vectors.”
Convolutional Neural Network Committees for Handwritten Character Classification (2011, methodology): “in the NIST SD 19 documentation [12], resulting in 37 distinct classes for this task.”