README: 22 datasets
Browse files
README.md
CHANGED
|
@@ -184,6 +184,18 @@ configs:
|
|
| 184 |
data_files:
|
| 185 |
- split: train
|
| 186 |
path: sidecar/ORIGA/*.parquet
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 187 |
- config_name: RAVIR
|
| 188 |
data_files:
|
| 189 |
- split: train
|
|
@@ -256,7 +268,7 @@ configs:
|
|
| 256 |
|
| 257 |
# OphGen — curated public ophthalmic datasets
|
| 258 |
|
| 259 |
-
Curated copies of
|
| 260 |
images as stored in the OphGen image store (de-identified, de-duplicated, native resolution), the train / validation / test
|
| 261 |
split used in the project, six captions per image (short / medium / long × English / Chinese) and all original task annotations.
|
| 262 |
|
|
@@ -292,13 +304,14 @@ In `all`, an image released by several datasets appears once per dataset (same `
|
|
| 292 |
| [Oculo](cards/Oculo.md) | Oculo | Conventional ocular B-scan ultrasonography | 1,189 | 162 | 245 | 2 |
|
| 293 |
| [ODIR](cards/ODIR.md) | ODIR-5K | CFP | 6,215 | 701 | 1,001 | 4 |
|
| 294 |
| [ORIGA](cards/ORIGA.md) | ORIGA | CFP | 293 | 32 | 325 | 4 |
|
|
|
|
| 295 |
| [RAVIR](cards/RAVIR.md) | RAVIR | IR/NIR | 20 | 2 | 19 | 2 |
|
| 296 |
| [REFUGE2](cards/REFUGE2.md) | REFUGE2 | CFP | 395 | 402 | 403 | 3 |
|
| 297 |
| [ROSSA](cards/ROSSA.md) | ROSSA | OCTA | 100 | 100 | 100 | 1 |
|
| 298 |
| [RVO-ME](cards/RVO-ME.md) | RVO-ME | OCT | 2,057 | 229 | 580 | 3 |
|
| 299 |
| [SJTU_Peripapillary](cards/SJTU_Peripapillary.md) | SJTU peripapillary OCT | OCT | 74 | 24 | 24 | 3 |
|
| 300 |
| [UWF-DR_2026](cards/UWF-DR_2026.md) | UWF-DR | UWF | 1,135 | 158 | 327 | 2 |
|
| 301 |
-
| **Total** | | |
|
| 302 |
|
| 303 |
Split rules: `official` = official split; `official_val_from_train` = 10% of the official training set (by patient) used as
|
| 304 |
validation; `labeled_7_1_2` = patient-level 7:1:2 split with a fixed seed (no official split); `pretrain_90_5_5` = 90:5:5 split
|
|
|
|
| 184 |
data_files:
|
| 185 |
- split: train
|
| 186 |
path: sidecar/ORIGA/*.parquet
|
| 187 |
+
- config_name: PALM
|
| 188 |
+
data_files:
|
| 189 |
+
- split: train
|
| 190 |
+
path: data/PALM/train-*.parquet
|
| 191 |
+
- split: validation
|
| 192 |
+
path: data/PALM/validation-*.parquet
|
| 193 |
+
- split: test
|
| 194 |
+
path: data/PALM/test-*.parquet
|
| 195 |
+
- config_name: PALM-sidecar
|
| 196 |
+
data_files:
|
| 197 |
+
- split: train
|
| 198 |
+
path: sidecar/PALM/*.parquet
|
| 199 |
- config_name: RAVIR
|
| 200 |
data_files:
|
| 201 |
- split: train
|
|
|
|
| 268 |
|
| 269 |
# OphGen — curated public ophthalmic datasets
|
| 270 |
|
| 271 |
+
Curated copies of 22 public ophthalmic imaging datasets, prepared for the OphGen ophthalmic generative-model project:
|
| 272 |
images as stored in the OphGen image store (de-identified, de-duplicated, native resolution), the train / validation / test
|
| 273 |
split used in the project, six captions per image (short / medium / long × English / Chinese) and all original task annotations.
|
| 274 |
|
|
|
|
| 304 |
| [Oculo](cards/Oculo.md) | Oculo | Conventional ocular B-scan ultrasonography | 1,189 | 162 | 245 | 2 |
|
| 305 |
| [ODIR](cards/ODIR.md) | ODIR-5K | CFP | 6,215 | 701 | 1,001 | 4 |
|
| 306 |
| [ORIGA](cards/ORIGA.md) | ORIGA | CFP | 293 | 32 | 325 | 4 |
|
| 307 |
+
| [PALM](cards/PALM.md) | PALM | CFP | 386 | 401 | 397 | 3 |
|
| 308 |
| [RAVIR](cards/RAVIR.md) | RAVIR | IR/NIR | 20 | 2 | 19 | 2 |
|
| 309 |
| [REFUGE2](cards/REFUGE2.md) | REFUGE2 | CFP | 395 | 402 | 403 | 3 |
|
| 310 |
| [ROSSA](cards/ROSSA.md) | ROSSA | OCTA | 100 | 100 | 100 | 1 |
|
| 311 |
| [RVO-ME](cards/RVO-ME.md) | RVO-ME | OCT | 2,057 | 229 | 580 | 3 |
|
| 312 |
| [SJTU_Peripapillary](cards/SJTU_Peripapillary.md) | SJTU peripapillary OCT | OCT | 74 | 24 | 24 | 3 |
|
| 313 |
| [UWF-DR_2026](cards/UWF-DR_2026.md) | UWF-DR | UWF | 1,135 | 158 | 327 | 2 |
|
| 314 |
+
| **Total** | | | 29,232 | 5,423 | 6,981 | |
|
| 315 |
|
| 316 |
Split rules: `official` = official split; `official_val_from_train` = 10% of the official training set (by patient) used as
|
| 317 |
validation; `labeled_7_1_2` = patient-level 7:1:2 split with a fixed seed (no official split); `pretrain_90_5_5` = 90:5:5 split
|