feat: DS4 and better infos
Browse files- ornithoscope.py +43 -13
ornithoscope.py
CHANGED
@@ -90,7 +90,6 @@ class OrnithoscopeConfig(datasets.BuilderConfig):
|
|
90 |
|
91 |
def __init__(
|
92 |
self,
|
93 |
-
classes: list[str],
|
94 |
train_json: str,
|
95 |
validation_json: str,
|
96 |
test_json: str,
|
@@ -106,11 +105,27 @@ class OrnithoscopeConfig(datasets.BuilderConfig):
|
|
106 |
**kwargs: keyword arguments forwarded to super.
|
107 |
"""
|
108 |
super().__init__(version=datasets.Version("1.0.0"), **kwargs)
|
109 |
-
self.classes =
|
|
|
|
|
|
|
|
|
110 |
self.train_json = train_json
|
111 |
self.validation_json = validation_json
|
112 |
self.test_json = test_json
|
113 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
114 |
|
115 |
class Ornithoscope(datasets.GeneratorBasedBuilder):
|
116 |
|
@@ -119,28 +134,43 @@ class Ornithoscope(datasets.GeneratorBasedBuilder):
|
|
119 |
OrnithoscopeConfig(
|
120 |
name="DS3",
|
121 |
description="The main dataset.",
|
122 |
-
classes=[], # TODO
|
123 |
train_json="sets/DS3_train.json",
|
124 |
validation_json="sets/DS3_val.json",
|
125 |
test_json="sets/DS3_test.json",
|
126 |
),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
127 |
]
|
128 |
|
129 |
def _info(self) -> datasets.DatasetInfo:
|
130 |
return datasets.DatasetInfo(
|
131 |
description=_DESCRIPTION,
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
138 |
homepage=_HOMEPAGE,
|
139 |
citation=_CITATION,
|
140 |
-
# task_templates=[ImageClassification(
|
141 |
-
# image_column="image",
|
142 |
-
# label_column="label",
|
143 |
-
# )],
|
144 |
)
|
145 |
|
146 |
def _split_generators(self, dl_manager: datasets.DownloadManager) -> list[datasets.SplitGenerator]:
|
|
|
90 |
|
91 |
def __init__(
|
92 |
self,
|
|
|
93 |
train_json: str,
|
94 |
validation_json: str,
|
95 |
test_json: str,
|
|
|
105 |
**kwargs: keyword arguments forwarded to super.
|
106 |
"""
|
107 |
super().__init__(version=datasets.Version("1.0.0"), **kwargs)
|
108 |
+
self.classes = self._get_classes([
|
109 |
+
train_json,
|
110 |
+
validation_json,
|
111 |
+
test_json,
|
112 |
+
])
|
113 |
self.train_json = train_json
|
114 |
self.validation_json = validation_json
|
115 |
self.test_json = test_json
|
116 |
|
117 |
+
def _get_classes(self, paths: list[str]) -> list[str]:
|
118 |
+
"""Get the classes from the paths."""
|
119 |
+
classes = []
|
120 |
+
for path in paths:
|
121 |
+
with open(path, 'r') as f:
|
122 |
+
file = json.load(f)
|
123 |
+
for value in file.values():
|
124 |
+
for box in value['boxes']:
|
125 |
+
if box['label'] not in classes:
|
126 |
+
classes.append(box['label'])
|
127 |
+
return classes
|
128 |
+
|
129 |
|
130 |
class Ornithoscope(datasets.GeneratorBasedBuilder):
|
131 |
|
|
|
134 |
OrnithoscopeConfig(
|
135 |
name="DS3",
|
136 |
description="The main dataset.",
|
|
|
137 |
train_json="sets/DS3_train.json",
|
138 |
validation_json="sets/DS3_val.json",
|
139 |
test_json="sets/DS3_test.json",
|
140 |
),
|
141 |
+
OrnithoscopeConfig(
|
142 |
+
name="DS4",
|
143 |
+
description="The new dataset.",
|
144 |
+
train_json="sets/DS4_train.json",
|
145 |
+
validation_json="sets/DS4_val.json",
|
146 |
+
test_json="sets/DS4_test.json",
|
147 |
+
),
|
148 |
]
|
149 |
|
150 |
def _info(self) -> datasets.DatasetInfo:
|
151 |
return datasets.DatasetInfo(
|
152 |
description=_DESCRIPTION,
|
153 |
+
features=datasets.Features(
|
154 |
+
{
|
155 |
+
"id_path": datasets.Value("string"),
|
156 |
+
"path": datasets.Value("string"),
|
157 |
+
"boxes": datasets.Sequence(
|
158 |
+
{
|
159 |
+
"label": datasets.Value("string"),
|
160 |
+
"xmin": datasets.Value("float32"),
|
161 |
+
"xmax": datasets.Value("float32"),
|
162 |
+
"ymin": datasets.Value("float32"),
|
163 |
+
"ymax": datasets.Value("float32"),
|
164 |
+
}
|
165 |
+
),
|
166 |
+
"size": {
|
167 |
+
"width": datasets.Value("int32"),
|
168 |
+
"height": datasets.Value("int32"),
|
169 |
+
},
|
170 |
+
},
|
171 |
+
),
|
172 |
homepage=_HOMEPAGE,
|
173 |
citation=_CITATION,
|
|
|
|
|
|
|
|
|
174 |
)
|
175 |
|
176 |
def _split_generators(self, dl_manager: datasets.DownloadManager) -> list[datasets.SplitGenerator]:
|