add img col

.gitignore

@@ -1,2 +1,4 @@
 rename.sh
 test.py
+*.wav
+*.jpg

chest_falsetto.py

@@ -9,7 +9,7 @@ _NAMES = ['m_chest', 'f_chest', 'm_falsetto', 'f_falsetto']
 
 _DBNAME = os.path.basename(__file__).split('.')[0]
 
-_HOMEPAGE = "https://huggingface.co/datasets/ccmusic-database/" + _DBNAME
+_HOMEPAGE = f"https://huggingface.co/datasets/ccmusic-database/{_DBNAME}"
 
 _CITATION = """\
 @dataset{zhaorui_liu_2021_5676893,
@@ -31,7 +31,7 @@ the Mel-spectrogram, MFCC, and spectral characteristics of each audio segment ar
 for a total of 5120 CSV files.
 """
 
-_URL = _HOMEPAGE + "/resolve/main/data/client_data.zip"
+_URL = f"{_HOMEPAGE}/resolve/main/data/client_data.zip"
 
 
 class chest_falsetto(datasets.GeneratorBasedBuilder):
@@ -41,6 +41,7 @@ class chest_falsetto(datasets.GeneratorBasedBuilder):
             features=datasets.Features(
                 {
                     "audio": datasets.Audio(sampling_rate=44_100),
+                    "image": datasets.Image(),
                     "label": datasets.features.ClassLabel(names=_NAMES),
                     "gender": datasets.features.ClassLabel(names=_GENDER),
                     "singing_method": datasets.features.ClassLabel(names=_CF),
@@ -63,50 +64,48 @@ class chest_falsetto(datasets.GeneratorBasedBuilder):
     def _split_generators(self, dl_manager):
         data_files = dl_manager.download_and_extract(_URL)
         files = dl_manager.iter_files([data_files])
-        trainset, validationset, testset = [], [], []
+        dataset = []
 
-        for _, path in enumerate(files):
-            trainset.append(path)
+        for path in files:
+            file_name = os.path.basename(path)
+            if file_name.endswith(".wav"):
+                sex = file_name.split('_')[1]
+                gender = 'male' if sex == 'm' else 'female'
+                method = file_name.split('_')[2][:-4]
+                dataset.append({
+                    "audio": path,
+                    "image": path.replace('.wav', '.jpg').replace('/audio/', '/image/'),
+                    "label": sex + '_' + method,
+                    "gender": gender,
+                    "singing_method": method,
+                })
 
-        random.shuffle(trainset)
-        data_count = len(trainset)
+        random.shuffle(dataset)
+        data_count = len(dataset)
         p80 = int(data_count * 0.8)
         p90 = int(data_count * 0.9)
-        validationset = trainset[p80:p90]
-        testset = trainset[p90:]
-        trainset = trainset[:p80]
 
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "files": trainset,
+                    "files": dataset[:p80],
                 },
             ),
             datasets.SplitGenerator(
                 name=datasets.Split.VALIDATION,
                 gen_kwargs={
-                    "files": validationset,
+                    "files": dataset[p80:p90],
                 },
             ),
             datasets.SplitGenerator(
                 name=datasets.Split.TEST,
                 gen_kwargs={
-                    "files": testset,
+                    "files": dataset[p90:],
                 },
             ),
         ]
 
     def _generate_examples(self, files):
         for i, path in enumerate(files):
-            file_name = os.path.basename(path)
-            if file_name.endswith(".wav"):
-                sex = file_name.split('_')[1]
-                gender = 'male' if sex == 'm' else 'female'
-                method = file_name.split('_')[2][:-4]
-                yield i, {
-                    "audio": path,
-                    "label": sex + '_' + method,
-                    "gender": gender,
-                    "singing_method": method,
-                }
+            yield i, path

data/client_data.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82a14d2259d441db6c84a9a4417468da6b5bcd89b3e95cb5c28e829fa12222f7
-size 39271503
+oid sha256:b477e7f813be6f1e6a4c47a86abb5d32c4b7789d1eb81553e5e716ee678d6d15
+size 70399735