Kaphathy commited on
Commit
b785ea7
·
verified ·
1 Parent(s): c55c6d8

README: 9 datasets

Browse files
Files changed (1) hide show
  1. README.md +13 -2
README.md CHANGED
@@ -24,6 +24,16 @@ configs:
24
  data_files:
25
  - split: train
26
  path: sidecar/DRIVE/*.parquet
 
 
 
 
 
 
 
 
 
 
27
  - config_name: GLEAM
28
  data_files:
29
  - split: train
@@ -108,7 +118,7 @@ configs:
108
 
109
  # OphGen — curated public ophthalmic datasets
110
 
111
- Curated copies of 8 public ophthalmic imaging datasets, prepared for the OphGen ophthalmic generative-model project:
112
  images as stored in the OphGen image store (de-identified, de-duplicated, native resolution), the train / validation / test
113
  split used in the project, six captions per image (short / medium / long × English / Chinese) and all original task annotations.
114
 
@@ -130,6 +140,7 @@ In `all`, an image released by several datasets appears once per dataset (same `
130
  | Config | Dataset | Modalities | Train | Validation | Test | Tasks |
131
  | :--- | :--- | :--- | ---: | ---: | ---: | ---: |
132
  | [DRIVE](cards/DRIVE.md) | DRIVE | CFP | 18 | 2 | 20 | 1 |
 
133
  | [GLEAM](cards/GLEAM.md) | GLEAM | IR/NIR, OCT | 1,436 | 480 | 480 | 1 |
134
  | [IDRiD](cards/IDRiD.md) | IDRiD (segmentation and grading parts) | CFP | 382 | 42 | 107 | 2 |
135
  | [RAVIR](cards/RAVIR.md) | RAVIR | IR/NIR | 20 | 2 | 19 | 2 |
@@ -137,7 +148,7 @@ In `all`, an image released by several datasets appears once per dataset (same `
137
  | [ROSSA](cards/ROSSA.md) | ROSSA | OCTA | 100 | 100 | 100 | 1 |
138
  | [RVO-ME](cards/RVO-ME.md) | RVO-ME | OCT | 2,057 | 229 | 580 | 3 |
139
  | [SJTU_Peripapillary](cards/SJTU_Peripapillary.md) | SJTU peripapillary OCT | OCT | 74 | 24 | 24 | 3 |
140
- | **Total** | | | 4,482 | 1,281 | 1,733 | |
141
 
142
  Split rules: `official` = official split; `official_val_from_train` = 10% of the official training set (by patient) used as
143
  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
 
24
  data_files:
25
  - split: train
26
  path: sidecar/DRIVE/*.parquet
27
+ - config_name: GAVE2
28
+ data_files:
29
+ - split: train
30
+ path: data/GAVE2/train-*.parquet
31
+ - split: validation
32
+ path: data/GAVE2/validation-*.parquet
33
+ - config_name: GAVE2-sidecar
34
+ data_files:
35
+ - split: train
36
+ path: sidecar/GAVE2/*.parquet
37
  - config_name: GLEAM
38
  data_files:
39
  - split: train
 
118
 
119
  # OphGen — curated public ophthalmic datasets
120
 
121
+ Curated copies of 9 public ophthalmic imaging datasets, prepared for the OphGen ophthalmic generative-model project:
122
  images as stored in the OphGen image store (de-identified, de-duplicated, native resolution), the train / validation / test
123
  split used in the project, six captions per image (short / medium / long × English / Chinese) and all original task annotations.
124
 
 
140
  | Config | Dataset | Modalities | Train | Validation | Test | Tasks |
141
  | :--- | :--- | :--- | ---: | ---: | ---: | ---: |
142
  | [DRIVE](cards/DRIVE.md) | DRIVE | CFP | 18 | 2 | 20 | 1 |
143
+ | [GAVE2](cards/GAVE2.md) | GAVE2 | FFA, CFP | 150 | 150 | 0 | 4 |
144
  | [GLEAM](cards/GLEAM.md) | GLEAM | IR/NIR, OCT | 1,436 | 480 | 480 | 1 |
145
  | [IDRiD](cards/IDRiD.md) | IDRiD (segmentation and grading parts) | CFP | 382 | 42 | 107 | 2 |
146
  | [RAVIR](cards/RAVIR.md) | RAVIR | IR/NIR | 20 | 2 | 19 | 2 |
 
148
  | [ROSSA](cards/ROSSA.md) | ROSSA | OCTA | 100 | 100 | 100 | 1 |
149
  | [RVO-ME](cards/RVO-ME.md) | RVO-ME | OCT | 2,057 | 229 | 580 | 3 |
150
  | [SJTU_Peripapillary](cards/SJTU_Peripapillary.md) | SJTU peripapillary OCT | OCT | 74 | 24 | 24 | 3 |
151
+ | **Total** | | | 4,632 | 1,431 | 1,733 | |
152
 
153
  Split rules: `official` = official split; `official_val_from_train` = 10% of the official training set (by patient) used as
154
  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