A Hybrid Approach and Unified Framework for Bibliographic Reference Extraction
Syed Tahseen Raza Rizvi, Andreas Dengel, Sheraz Ahmed

TL;DR
This paper introduces DeepBiRD, a novel layout-based method for extracting references from scientific publications, supported by a large annotated dataset and integrated into a flexible framework, achieving high accuracy and generalization across styles.
Contribution
It presents a new layout-driven reference detection method, a large annotated dataset, and a unified framework for robust bibliographic reference extraction.
Findings
DeepBiRD achieved 98.56% AP50 on the new dataset.
Outperformed existing state-of-the-art methods.
Demonstrated strong generalization across datasets and styles.
Abstract
Publications are an integral part in a scientific community. Bibliographic reference extraction from scientific publication is a challenging task due to diversity in referencing styles and document layout. Existing methods perform sufficiently on one dataset however, applying these solutions to a different dataset proves to be challenging. Therefore, a generic solution was anticipated which could overcome the limitations of the previous approaches. The contribution of this paper is three-fold. First, it presents a novel approach called DeepBiRD which is inspired by human visual perception and exploits layout features to identify individual references in a scientific publication. Second, we release a large dataset for image-based reference detection with 2401 scans containing 38863 references, all manually annotated for individual reference. Third, we present a unified and highly…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
MethodsRegion Proposal Network · Softmax · Convolution · RoIAlign · Mask R-CNN
