Similarity-based Text Recognition by Deeply Supervised Siamese Network

TL;DR

This paper introduces a deep supervision Siamese network for text recognition that measures visual similarity to predict unlabeled texts, significantly reducing human effort and outperforming traditional methods.

Contribution

The paper presents a novel deeply supervised Siamese network architecture for text recognition that improves similarity measurement and label prediction accuracy.

Findings

Reduces human labeling effort by 50-85%

Achieves less than 0.5% error rate

Outperforms conventional Siamese networks in accuracy

Abstract

In this paper, we propose a new text recognition model based on measuring the visual similarity of text and predicting the content of unlabeled texts. First a Siamese convolutional network is trained with deep supervision on a labeled training dataset. This network projects texts into a similarity manifold. The Deeply Supervised Siamese network learns visual similarity of texts. Then a K-nearest neighbor classifier is used to predict unlabeled text based on similarity distance to labeled texts. The performance of the model is evaluated on three datasets of machine-print and hand-written text combined. We demonstrate that the model reduces the cost of human estimation by $50\%-85\%$. The error of the system is less than $0.5\%$. The proposed model outperform conventional Siamese network by finding visually-similar barely-readable and readable text, e.g. machine-printed, handwritten, due to deep supervision. The results also demonstrate that the predicted labels are sometimes better than human labels e.g. spelling correction.