Pattern Spotting in Historical Documents Using Convolutional Models
This work addresses a domain-specific challenge in historical document analysis, offering incremental improvements in efficiency and accuracy for researchers and archivists.
The paper tackled the problem of pattern spotting in historical document images without prior object information by proposing a convolutional neural network approach using RetinaNet for feature extraction, achieving better pattern localization and reduced storage requirements on the DocExplore dataset, though it struggled with retrieving pages containing multiple query instances.
Pattern spotting consists of searching in a collection of historical document images for occurrences of a graphical object using an image query. Contrary to object detection, no prior information nor predefined class is given about the query so training a model of the object is not feasible. In this paper, a convolutional neural network approach is proposed to tackle this problem. We use RetinaNet as a feature extractor to obtain multiscale embeddings of the regions of the documents and also for the queries. Experiments conducted on the DocExplore dataset show that our proposal is better at locating patterns and requires less storage for indexing images than the state-of-the-art system, but fails at retrieving some pages containing multiple instances of the query.