DoB24/fundus-10class-augmented

Fundus 10-Class Augmented Dataset

Augmented and split-curated version of the Mendeley fundus dataset, used in the Master's thesis "Classification of Fundus Lesion Images Using Deep Learning Models" (Xidian University, 2026).

Models trained on this dataset: DoB24/fundus-9model-benchmark.

Summary

• 10 disease classes (see table below).
• 5,335 original images; 16,242 augmented images (rotation, flip, brightness/contrast jitter, Gaussian blur).
• Group-aware test split (splits/holdout_split_augmented.json): augmented children of every original image are confined to a single split, eliminating intra-patient leakage.
• Image size: variable (resized to 224×224 / 299×299 during training).

Class distribution

#	Class	Original	Augmented
1	Central Serous Chorioretinopathy [Color Fundus]	101	606
2	Diabetic Retinopathy	1,509	3,444
3	Disc Edema	127	762
4	Glaucoma	1,349	2,880
5	Healthy	1,024	2,676
6	Macular Scar	444	1,937
7	Myopia	500	2,251
8	Pterygium	17	102
9	Retinal Detachment	125	750
10	Retinitis Pigmentosa	139	834
—	Total	5,335	16,242

File layout

images/augmented/<class_name>/<file>.jpg     # 16K augmented imgs
images/original/<class_name>/<file>.jpg      #  5K original imgs
splits/holdout_split_augmented.json          # pHash-grouped 5-fold manifest

Group-aware splitting

We computed a 64-bit perceptual hash (pHash) per image and linked each augmented image to its nearest original at Hamming distance ≤ 8 to define a group_id. A StratifiedGroupKFold (k=5, fold-0 reported) produced 15 068 train / 3 301 val / 3 208 test images.

Loading

from datasets import load_dataset
ds = load_dataset("DoB24/fundus-10class-augmented")

Or download manually:

from huggingface_hub import snapshot_download
snapshot_download("DoB24/fundus-10class-augmented", repo_type="dataset", local_dir="./fundus_data")

Source & license

Original images: Mendeley dataset DOI 10.17632/s9bfhswzjb.1 — CC BY 4.0. Augmented derivatives inherit CC BY 4.0.

Citation

@dataset{nayan2023fundus,
  title  = {A Curated Dataset of Retinal Fundus Images for Disease Classification},
  year   = {2023},
  doi    = {10.17632/s9bfhswzjb.1},
  url    = {https://data.mendeley.com/datasets/s9bfhswzjb/1}
}

Original (Unaugmented) Dataset

The 5,335 source images (before augmentation) are provided as a single gzipped tar archive for fast download:

• File: original_dataset.tar.gz (1.6 GB)
• Contents: Original_Dataset/<class>/*.{jpg,png} — 10 classes, 5,335 images total
• Download & extract:

huggingface-cli download DoB24/fundus-10class-augmented original_dataset.tar.gz --repo-type dataset --local-dir .
tar xzf original_dataset.tar.gz

The augmented split (16,242 images under images/augmented/) is the one used for all reported training results. The original archive is provided for full reproducibility of the augmentation pipeline (see code/ in the model repo DoB24/fundus-9model-benchmark).