Learning a Similarity Metric Discriminatively with Application to Ancient Character Recognition

Abstract

The process of learning good representation in deep learning may prove difficult when the data is insufficient. In this paper, we propose a Siamese similarity network for one-shot ancient character recognition based on a similarity learning method to directly learn input similarity, and then use the trained model to establish one shot classification task for recognition. Multi-scale fusion backbone structure and embedded structure are proposed in the network to improve the model's ability to extract features. we also propose the soft similarity contrast loss function for the first time. It ensures the optimization of similar images with higher similarity and different classes of images with greater differences while reducing the over-optimization of back-propagation leading to model overfitting. A large number of experiments show that our proposed method has achieved high-efficiency discriminative performance, and obtained the best performance over the methods of traditional deep learning and other classic one-shot learning.