Datasets:
Formats:
parquet
Languages:
English
Size:
100K - 1M
ArXiv:
Tags:
query-by-example-spoken-term-detection
audio-slot-filling
speaker-diarization
automatic-speaker-verification
License:
Update superb.py
Browse files
superb.py
CHANGED
|
@@ -435,14 +435,13 @@ class Superb(datasets.GeneratorBasedBuilder):
|
|
| 435 |
label = "_unknown_"
|
| 436 |
yield key, {"file": audio_file, "audio": audio_file, "label": label}
|
| 437 |
elif self.config.name == "ic":
|
| 438 |
-
|
| 439 |
-
csv_path = os.path.join(root_path, "data", f"{split}_data.csv")
|
| 440 |
with open(csv_path, encoding="utf-8") as csv_file:
|
| 441 |
csv_reader = csv.reader(csv_file, delimiter=",", skipinitialspace=True)
|
| 442 |
next(csv_reader)
|
| 443 |
for row in csv_reader:
|
| 444 |
key, file_path, speaker_id, text, action, object_, location = row
|
| 445 |
-
audio_path = os.path.join(
|
| 446 |
yield key, {
|
| 447 |
"file": audio_path,
|
| 448 |
"audio": audio_path,
|
|
|
|
| 435 |
label = "_unknown_"
|
| 436 |
yield key, {"file": audio_file, "audio": audio_file, "label": label}
|
| 437 |
elif self.config.name == "ic":
|
| 438 |
+
csv_path = os.path.join(archive_path, "data", f"{split}_data.csv")
|
|
|
|
| 439 |
with open(csv_path, encoding="utf-8") as csv_file:
|
| 440 |
csv_reader = csv.reader(csv_file, delimiter=",", skipinitialspace=True)
|
| 441 |
next(csv_reader)
|
| 442 |
for row in csv_reader:
|
| 443 |
key, file_path, speaker_id, text, action, object_, location = row
|
| 444 |
+
audio_path = os.path.join(archive_path, file_path)
|
| 445 |
yield key, {
|
| 446 |
"file": audio_path,
|
| 447 |
"audio": audio_path,
|