A Survey of Handwritten Character Recognition with MNIST and EMNIST

Abstract

This paper summarizes the top state-of-the-art contributions reported on the MNIST dataset for handwritten digit recognition. This dataset has been extensively used to validate novel techniques in computer vision, and in recent years, many authors have explored the performance of convolutional neural networks (CNNs) and other deep learning techniques over this dataset. To the best of our knowledge, this paper is the first exhaustive and updated review of this dataset; there are some online rankings, but they are outdated, and most published papers survey only closely related works, omitting most of the literature. This paper makes a distinction between those works using some kind of data augmentation and works using the original dataset out-of-the-box. Also, works using CNNs are reported separately; as they are becoming the state-of-the-art approach for solving this problem. Nowadays, a significant amount of works have attained a test error rate smaller than 1% on this dataset; which is becoming non-challenging. By mid-2017, a new dataset was introduced: EMNIST, which involves both digits and letters, with a larger amount of data acquired from a database different than MNIST's. In this paper, EMNIST is explained and some results are surveyed.

Convolutional Extreme Learning Machines: A Systematic Review

“…Dataset [64] Classification of digestive organs disease Own dataset [65] Liver tumor classification Elazig University Hospital [66] White blood cell detection BCCD dataset [67] Histopathological image classification ADL dataset [68] Cerebral microbleed diagnosis Own dataset [69] Cervical cancer classification Herlev dataset [61] Brain tumor classification CGA-GBM database [70] Micro-nodules classification LIDC/IDRI dataset [62] Brain tumor classification Brain T1-weighed CE-MRI dataset [63] Brain tumor classification Brain tumor MRI dataset [71] Classification of anomalies in the human retina Duke and HUCM datasets [72] Hepatocellular carcinoma classification ICPR 2014 HEp-2 cell dataset [73]. Several works proposed digit and character recognition for applications such as handwriting recognition [73]. Handwritten digit or character recognition can be applied to several tasks: text categorization from images, classification of documents, signature recognition, etc.…”

Survey on Segmentation and Recognition of Handwritten Arabic Script

“…A few problems related with Urdu handwritten document make the mission of improving the framework for handwritten recognition more complicated and challenging. A number of these problems are variances within style of writing, even from a similar writer, illegible handwriting, poor quality or degradation of image because of script's cursive nature [49].…”

Combination of Convolutional Neural Network Architecture and its Learning Method for Rotation‐Invariant Handwritten Digit Recognition

“…We use a handwritten digit dataset called MNIST with 60 000 training images and 10 000 testing images [1,7]. These images [7][8][9].…”

“…We use a handwritten digit dataset called MNIST with 60 000 training images and 10 000 testing images [1,7]. These images are augmented four times with rotation angles of 0 ∘ , 90 ∘ , 180 ∘ , and 270 ∘ .…”

“…Handwritten digit recognition is a classical and important task in computer vision. Recently, convolutional neural networks (CNNs) have achieved state-of-the-art results in digit recognition [1]. However, standard CNNs are not invariant to a large rotation transformation of input images.…”