/Bengali-Handwritten-Character-Recognition-Using-Convolutional-Neural-Networks

Implementation of convolutional neural networks (CNNs) of increasing complexity for classification of handwritten Bengali characters. The deepest network achieves 94.2% accuracy on test data, compared to 85.96% reported on the latest paper using CNNs for Bengali OCR.

Primary LanguageJupyter NotebookMIT LicenseMIT

Issues