Classification of Hand-Written Digits using Deep-Learning Technique
Marriam Nawaz, Department of Software Engineering, University of Engineering and Technology, Taxila, Pakistan.
Ali Javed, Department of Software Engineering, University of Engineering and Technology, Taxila, Pakistan.
Tahira Nazir, Department of Computing, Riphah International University, Islamabad, Pakistan.
Momina Masood, Department of Computer Science, University of Engineering and Technology, Taxila, Pakistan.
Awais Mehmood, Department of Computer Science, University of Engineering and Technology, Taxila, Pakistan.
Faheem Saleem, Department of Computer Science, University of Engineering and Technology, Taxila, Pakistan.
Corresponding Author:
Tahira Nazir (tahira.nazir@riphah.edu.pk)
Abstract:
In the field of information processing, handwritten digit recognition (HDR) plays an important role as an application. However, this process becomes complex and tedious due to the enormous variations of writing styles. In this paper, we have proposed the deep-learning-based model which is Fast-RCNN for localization and classification of numeral images. Firstly, we have annotated the input numeral images, which can be localized through Fast-RCNN and then classify into ten categories 0 to 9. The challenging dataset “MINIST” is employed for the evaluation of our model. The experimental outcomes reveal that the presented technique can precisely and accurately classify the numeral images in the presence of different writing style variations, and under the existence of noise, blurring, etc. in input images.
Keywords:
Handwritten; Digit; Fast-RCNN; Deep Learning; Classification