GIRAFE: Glottal Imaging Dataset for Advanced Segmentation, Analysis, and Facilitative Playbacks Evaluation

TL;DR

GIRAFE addresses the lack of publicly available, richly annotated HSV datasets for glottal-gap segmentation by providing $65$ HSV recordings from $50$ subjects (including healthy and disordered cases) with manual segmentation masks and FP gold standards. The dataset combines color HSV video, comprehensive metadata, manual and automatic segmentation results (InP and Loh), and multiple Facilitative Playback representations (GAW, GVG, PVG), along with baseline DL and traditional methods evaluated on a consistent split. Key contributions include the multi-faceted data organization (Raw_Data, Seg_FP-Results, Training), detailed FP generation, and accessible code resources (GitHub, Seg_FP-Results notebook, MATLAB scripts) to support training, evaluation, and reproducibility. This open resource is poised to accelerate robust glottal-gap segmentation methods, FP development, and cross-dataset generalization, complementing existing datasets such as BAGLS and enabling more reproducible clinical tooling.

Abstract

The advances in the development of Facilitative Playbacks extracted from High-Speed videoendoscopic sequences of the vocal folds are hindered by a notable lack of publicly available datasets annotated with the semantic segmentations corresponding to the area of the glottal gap. This fact also limits the reproducibility and further exploration of existing research in this field. To address this gap, GIRAFE is a data repository designed to facilitate the development of advanced techniques for the semantic segmentation, analysis, and fast evaluation of High-Speed videoendoscopic sequences of the vocal folds. The repository includes 65 high-speed videoendoscopic recordings from a cohort of 50 patients (30 female, 20 male). The dataset comprises 15 recordings from healthy controls, 26 from patients with diagnosed voice disorders, and 24 with an unknown health condition. All of them were manually annotated by an expert, including the masks corresponding to the semantic segmentation of the glottal gap. The repository is also complemented with the automatic segmentation of the glottal area using different state-of-the-art approaches. This data set has already supported several studies, which demonstrates its usefulness for the development of new glottal gap segmentation algorithms from High-Speed-Videoendoscopic sequences to improve or create new Facilitative Playbacks. Despite these advances and others in the field, the broader challenge of performing an accurate and completely automatic semantic segmentation method of the glottal area remains open.

Figures (8)

• Figure 1: Illustration of a complete glottal cycle extracted from laryngeal HSV, along with three FP synthesized from them: GVG, PVG and GAW.
• Figure 2: Workflow for generating the GIRAFE dataset. Participants are of varying ages, gender, and health conditions, and were recruited at the Otorhinolaryngology Service of Hospital General Gregorio Marañón in Madrid. Glottal segmentation was conducted using manual, automatic, and semi-automatic techniques. To validate the GIRAFE dataset, two deep neural networks were trained, and the resulting segmentations were found to be highly consistent with the expert manual annotations.
• Figure 3: Age and sex distribution of the recordings in the GIRAFE dataset.
• Figure 4: Illustration of four cases with voice disorders: (a) cervicotomy for cervical hernia with postoperative diplophonia; (b) multinodular goiter; (c) vagal paraganglioma with fat infiltration; and, (d) vocal fold polyp.
• Figure 5: Database tree structure: hierarchical representation of the data records and their organization within the database. For simplicity, the contents of only one patient are displayed for each subfolder.