Update Populus_Stomatal_Images_Datasets.py
Browse files
Populus_Stomatal_Images_Datasets.py
CHANGED
@@ -109,30 +109,37 @@ class NewDataset(datasets.GeneratorBasedBuilder):
|
|
109 |
)
|
110 |
|
111 |
def _split_generators(self, dl_manager):
|
112 |
-
#
|
113 |
data_files = dl_manager.download_and_extract({
|
114 |
"csv": "https://huggingface.co/datasets/XintongHe/Populus_Stomatal_Images_Datasets/resolve/main/data/Labeled Stomatal Images.csv",
|
115 |
-
"
|
|
|
116 |
})
|
117 |
-
extracted_images_path = 'https://huggingface.co/datasets/XintongHe/Populus_Stomatal_Images_Datasets/resolve/main/data/Labeled Stomatal Images_config.zip'
|
118 |
-
|
119 |
-
# Load the CSV file containing species and scientific names
|
120 |
-
species_info = pd.read_csv(os.path.join(extracted_images_path, 'Labeled Stomatal Images.csv'))
|
121 |
-
|
122 |
-
# Get all image filenames from the new configuration directory
|
123 |
-
all_image_filenames = [f for f in os.listdir(extracted_images_path) if f.endswith('.jpg')]
|
124 |
-
|
125 |
|
|
|
|
|
126 |
|
127 |
-
#
|
128 |
-
|
129 |
-
|
130 |
-
|
131 |
-
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
)]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
136 |
|
137 |
|
138 |
def _parse_yolo_labels(self, label_path, width, height):
|
|
|
109 |
)
|
110 |
|
111 |
def _split_generators(self, dl_manager):
|
112 |
+
# Download the CSV file and both ZIP files
|
113 |
data_files = dl_manager.download_and_extract({
|
114 |
"csv": "https://huggingface.co/datasets/XintongHe/Populus_Stomatal_Images_Datasets/resolve/main/data/Labeled Stomatal Images.csv",
|
115 |
+
"zip_all": "https://huggingface.co/datasets/XintongHe/Populus_Stomatal_Images_Datasets/resolve/main/data/Labeled Stomatal Images.zip",
|
116 |
+
"zip_config": "https://huggingface.co/datasets/XintongHe/Populus_Stomatal_Images_Datasets/resolve/main/data/Labeled Stomatal Images_config.zip"
|
117 |
})
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
118 |
|
119 |
+
# Load the CSV file containing species and scientific names
|
120 |
+
species_info_full = pd.read_csv(data_files["csv"])
|
121 |
|
122 |
+
# Assume the config ZIP is extracted to 'Labeled Stomatal Images_config' folder
|
123 |
+
extracted_config_path = os.path.join(dl_manager._cache_dir, "Labeled Stomatal Images_config")
|
124 |
+
|
125 |
+
# List all image filenames in the extracted config directory
|
126 |
+
all_image_filenames_config = [f for f in os.listdir(extracted_config_path) if f.endswith('.jpg')]
|
127 |
+
|
128 |
+
# Filter the species_info dataframe to only include the rows with filenames that are in the new configuration
|
129 |
+
# Here, we're extracting the filename without the extension to match with the 'FileName' column in the CSV
|
130 |
+
species_info = species_info_full[species_info_full['FileName'].isin([os.path.splitext(f)[0] for f in all_image_filenames_config])]
|
131 |
+
|
132 |
+
return [
|
133 |
+
datasets.SplitGenerator(
|
134 |
+
name=datasets.Split.TRAIN,
|
135 |
+
gen_kwargs={
|
136 |
+
"filepaths": all_image_filenames_config,
|
137 |
+
"species_info": species_info,
|
138 |
+
"data_dir": extracted_config_path
|
139 |
+
},
|
140 |
+
)
|
141 |
+
]
|
142 |
+
|
143 |
|
144 |
|
145 |
def _parse_yolo_labels(self, label_path, width, height):
|