Performance Evaluation of Advanced Deep Learning Architectures for Offline Handwritten Character Recognition
This paper presents a hand-written character recognition comparison and performance evaluation for robust and precise classification of different hand-written characters. The system utilizes advanced multilayer deep neural network by collecting features from raw pixel values. The hidden layers stack deep hierarchies of non-linear features since learning complex features from conventional neural networks is very challenging. Two state of the art deep learning architectures were used which includes Caffe AlexNet and GoogleNet models in NVIDIA DIGITS.The frameworks were trained and tested on two different datasets for incorporating diversity and complexity. One of them is the publicly available dataset i.e. Chars74K comprising of 7705 characters and has upper and lowercase English alphabets, along with numerical digits. While the other dataset created locally consists of 4320 characters. The local dataset consists of 62 classes and was created by 40 subjects. It also consists upper and lowercase English alphabets, along with numerical digits. The overall dataset is divided in the ratio of 80% for training and 20% for testing phase. The time required for training phase is approximately 90 minutes. For validation part, the results obtained were compared with the groundtruth. The accuracy level achieved with AlexNet was 77.77% and 88.89% with Google Net. The higher accuracy level of GoogleNet is due to its unique combination of inception modules, each including pooling, convolutions at various scales and concatenation procedures.