OLIVES: Ophthalmic Labels for Investigating Visual Eye Semantics

Longitudinal OCT + fundus dataset from AMD/DME patients across multiple clinical visits. 9,408 OCT B-scans have biomarker labels for 8 categories (fluid, drusen, scarring, PED variants, etc.); 78,000+ images include clinical measurements (BCVA, CST). Also includes 1,268 paired fundus photographs.

At a glance

Field	Value
Short name	`olives`
Full name	OLIVES: Ophthalmic Labels for Investigating Visual Eye Semantics
Primary category	`oct`
Contained modalities	oct, fundus, tabular
Tasks	multilabel, classification
Samples	9,408
Classes	8 (Fluid, Hard Exudate, Soft Exudate, Drusen, Scar, Laser Scar, Fibrovascular PED, Serous PED)
Splits	train, val, test
Size	20.0 GB
Source-stated terms	CC BY 4.0
Normalized terms	`cc-by`
Descriptive screening label	Standard label without an explicit NC clause; not a permission finding
Terms scope	`dataset_files`
Access friction	`anonymous_direct`
Route backend	Zenodo
Availability	`available` (checked 2026-07-21)
Acquisition support	`standard_platform_supported`
Legacy sample-loader status	Standard loader included

Notes

9,408 biomarker-labelled B-scans from 96 patients (6 clinical visits × 49 B-scans/visit). Full OCT volume set ~62,000 B-scans. Also available on HuggingFace: gOLIVES/OLIVES_Dataset. Contains paired fundus photos and clinical metadata.

Access preflight and acquisition

CLI
Python

# Read-only preflight
eyehub download olives --data-dir ./data --dry-run --json

# Explicit transfer, only when preflight reports supported behavior
eyehub download olives --data-dir ./data

from eyedatahub.acquisition import preflight_dataset
from eyedatahub.datasets.registry import REGISTRY

ds = REGISTRY.get_dataset('olives')
print(preflight_dataset(ds, './data'))  # no transfer

Upstream page: zenodo.org/records

Source-term evidence: zenodo.org/records

Loader example

This entry includes a standard DatasetSample loader.

from pathlib import Path
from eyedatahub.datasets.registry import REGISTRY

data_dir = Path('~/.eyedatahub/data').expanduser()
ds = REGISTRY.get_dataset('olives')
samples = ds.load(data_dir, split='test')
for s in samples[:5]:
    print(s.sample_id, s.label, s.image_path)

Citation

BibTeX
Plain text

@misc{olives,
  title  = { OLIVES: Ophthalmic Labels for Investigating Visual Eye Semantics },
  note   = { Prabhushankar M et al., 'OLIVES Dataset: Ophthalmic Labels for Investigating Visual Eye Semantics', NeurIPS Datasets & Benchmarks 2022. https://zenodo.org/records/7105232 },
  year   = { 2022 },
  url    = { https://zenodo.org/records/7105232 },
}

Prabhushankar M et al., 'OLIVES Dataset: Ophthalmic Labels for Investigating Visual Eye Semantics', NeurIPS Datasets & Benchmarks 2022. https://zenodo.org/records/7105232

Source-stated terms

Raw source string: CC BY 4.0
Normalized category: cc-by
Apparent scope: dataset_files
Descriptive screening label: Standard label without an explicit NC clause; not a permission finding

⚠️ Source-stated terms, scope, and normalized labels are curation metadata, not legal advice or a permission finding. Review the current official source before transfer or reuse.

lmod_plus: LMOD+ Multimodal Ophthalmology Benchmark (32,633 records, unknown)
grape: GRAPE: Glaucoma Real-world Appraisal Progression Ensemble (1,115 records, cc0)
eyecare_100k: Eyecare-100K: Multimodal Ophthalmology VQA Corpus (102,000 records, unknown)
multieye: MultiEYE: OCT-Enhanced Fundus Multi-Disease Benchmark (58,036 records, mit)
harvard_fairvision: Harvard-FairVision (AMD + DR + Glaucoma, paired SLO + OCT) (30,000 records, cc-by-nc-nd)
mmrdr: MMRDR: Multi-Modal Retinal Diabetic Retinopathy Dataset (24,460 records, cc-by)
x_pcr: X-PCR Ophthalmology Progressive Clinical Reasoning Benchmark (18,700 records, unknown)
fprm_retina: FPRM Multimodal Eye Imaging and Psychological Assessment Dataset (3,361 records, research-only)

At a glance​

Notes​

Access preflight and acquisition​

Loader example​

Citation​

Source-stated terms​

Related datasets with shared modalities​