iDocV2: Leveraging Self-Supervision and Open-Set Detection for Improving Pattern Spotting in Historical Documents

arXiv cs.CV / 4/21/2026

📰 NewsDeveloper Stack & InfrastructureModels & Research

共有:

Key Points

The paper proposes iDocV2, a new pattern-spotting model for historical documents that combines a better encoder (iDoc) with self-supervised training and an open-set detector.
It targets limitations of existing state-of-the-art approaches, which still fall short in both precision (especially for small non-square queries) and runtime (up to ~7 seconds per search in the DocExplore dataset).
iDocV2 is reported to achieve competitive performance for both pattern spotting and document retrieval while improving search speed by about 10x.
For the previously weak case of small non-square queries, the model reaches a new state-of-the-art precision of 0.612, improving over prior results.
Compared with the previous version, iDocV2 uses non-maximum suppression to reduce false positives.

Abstract

Considering the imminent massification of digital books, it has become critical to facilitate searching collections through graphical patterns. Current strategies for document retrieval and pattern spotting in historical documents still need to be improved. State-of-the-art strategies achieve an overall precision of

0.494

for pattern spotting, where the precision for small non-square queries reaches 0.427. In addition, the processing time is excessive, requiring up to 7 seconds for searching in the DocExplore dataset due to a dense-based strategy used by SOTA models. Therefore, we propose a new model based on a better encoder (iDoc), trained under a self-supervised strategy, and an open-set detector to accelerate searching. Our model achieves competitive results with state-of-the-art pattern spotting and document retrieval, improving speed by 10x. Furthermore, our model reaches a new SOTA performance on the small non-square queries, achieving a new precision of 0.612.Different from the previous version, this leverages non-maximum suppression to reduce false positives.