Handwritten Chinese Character Recognition by Convolutional Neural Network and Similarity Ranking
This work addresses recognition accuracy for users in document digitization or OCR applications, but it is incremental as it builds on existing CNN methods by modifying the loss function.
The paper tackled the problem of handwritten Chinese character recognition by proposing a combination of cross-entropy with similarity ranking functions as the loss function in convolutional neural networks, resulting in higher accuracy, with SoftMax cross-entropy combined with Average variance similarity achieving the highest accuracy.
Convolution Neural Networks (CNN) have recently achieved state-of-the art performance on handwritten Chinese character recognition (HCCR). However, most of CNN models employ the SoftMax activation function and minimize cross entropy loss, which may cause loss of inter-class information. To cope with this problem, we propose to combine cross entropy with similarity ranking function and use it as loss function. The experiments results show that the combination loss functions produce higher accuracy in HCCR. This report briefly reviews cross entropy loss function, a typical similarity ranking function: Euclidean distance, and also propose a new similarity ranking function: Average variance similarity. Experiments are done to compare the performances of a CNN model with three different loss functions. In the end, SoftMax cross entropy with Average variance similarity produce the highest accuracy on handwritten Chinese characters recognition.