pedroodb/glosl-elementary23

GloSL — Greek Elementary Sign Language Dataset (GSSL) → Greek

Part of GloSL, a set of continuous sign language translation corpora in one format.

• Sign language: Greek Sign Language (GSL/ENG) (gss)
• Spoken language: Greek (el)
• Domain: primary-school educational content, studio recordings
• Samples: 29,652 (72.79 h) — splits {'train': 23701, 'dev': 3207, 'test': 2744}
• Split policy: glosl-splits-v1, glosl — GloSL-defined (keep_existing): GloSL v0.2: deterministic signer-disjoint; the paper's curated SLT subset (7,348/512/512) is not in the Zenodo release (see card.json and diagnostics/)
• License: CC BY 4.0 — public; source: https://zenodo.org/records/7847052
• Cite: Chatzis, Kourbetis, Paramythiotis et al., Greek Elementary Sign Language Dataset, Zenodo 2023, doi:10.5281/zenodo.7847052

Files

annotations.csv, card.json, poses/<NNNN>/<id>.pose, videos/<NNNN>/<id>.mp4 (same frames as the pose). card.json holds the full provenance.

annotations.csv columns (same in every GloSL dataset; empty when not applicable): id, split, text, gloss, gloss_type, signer, start, end, fps, n_frames, video_ref, pose_file, video_file.

Poses

mediapipe-holistic, 543 points (POSE_LANDMARKS 33, FACE_LANDMARKS 468, LEFT_HAND_LANDMARKS 21, RIGHT_HAND_LANDMARKS 21), 3D. Coordinates: pixels of the frame the pose refers to (header width x height, depth 0); z raw. Confidence: body = MediaPipe visibility (1.0 where the source shipped none); face/hands 1 when detected; missing point = (0,0,0) with confidence 0. Frame rate: native, per file (also in annotations.csv). Normalization (glosl-norm-v1) is applied at load time, not stored. Note: MediaPipe Holistic extracted from the single-signer studio clips (no YOLO crop); native fps; normalization applied at LOAD time (KB/05).