AKHCRNet: Bengali Handwritten Character Recognition Using Deep Learning

29 Aug 2020  ·  Akash Roy ·

I propose a state of the art deep neural architectural solution for handwritten character recognition for Bengali alphabets, compound characters as well as numerical digits that achieves state-of-the-art accuracy 96.8% in just 11 epochs. Similar work has been done before by Chatterjee, Swagato, et al. but they achieved 96.12% accuracy in about 47 epochs. The deep neural architecture used in that paper was fairly large considering the inclusion of the weights of the ResNet 50 model which is a 50 layer Residual Network. This proposed model achieves higher accuracy as compared to any previous work & in a little number of epochs. ResNet50 is a good model trained on the ImageNet dataset, but I propose an HCR network that is trained from the scratch on Bengali characters without the "Ensemble Learning" that can outperform previous architectures.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Handwriting Recognition BanglaLekha Isolated Dataset AKHCRNet Accuracy 96.8 # 1
Epochs 11 # 1
Cross Entropy Loss 0.21612 # 1
Transfer Learning BanglaLekha Isolated Dataset Chatterjee, Dutta et al.[1] Transfer Learning on ResNet-50 91.13 % (50 Char) + 98.42% (Numbers) Accuracy 96.12 # 1