Image Retrieval and Pattern Spotting using Siamese Neural Network

TL;DR

This paper introduces a Siamese Neural Network-based method for image retrieval and pattern spotting in document collections, achieving high accuracy without manual feature engineering.

Contribution

The paper proposes a novel similarity-based representation learned via Siamese Neural Networks for document image retrieval and pattern spotting.

Findings

Achieves 0.94 mAP for retrieval

Achieves 0.83 mAP for pattern spotting

Demonstrates robustness across different feature map sizes

Abstract

This paper presents a novel approach for image retrieval and pattern spotting in document image collections. The manual feature engineering is avoided by learning a similarity-based representation using a Siamese Neural Network trained on a previously prepared subset of image pairs from the ImageNet dataset. The learned representation is used to provide the similarity-based feature maps used to find relevant image candidates in the data collection given an image query. A robust experimental protocol based on the public Tobacco800 document image collection shows that the proposed method compares favorably against state-of-the-art document image retrieval methods, reaching 0.94 and 0.83 of mean average precision (mAP) for retrieval and pattern spotting (IoU=0.7), respectively. Besides, we have evaluated the proposed method considering feature maps of different sizes, showing the impact of reducing the number of features in the retrieval performance and time-consuming.