Upload struct_amb_ind.py with huggingface_hub
Browse files- struct_amb_ind.py +174 -0
struct_amb_ind.py
ADDED
@@ -0,0 +1,174 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
from itertools import chain
|
3 |
+
from pathlib import Path
|
4 |
+
from typing import Dict, List, Tuple
|
5 |
+
|
6 |
+
import datasets
|
7 |
+
|
8 |
+
from seacrowd.utils import schemas
|
9 |
+
from seacrowd.utils.configs import SEACrowdConfig
|
10 |
+
from seacrowd.utils.constants import Licenses, Tasks
|
11 |
+
|
12 |
+
_CITATION = """\
|
13 |
+
@inproceedings{widiaputri-etal-5641,
|
14 |
+
author = {Widiaputri, Ruhiyah Faradishi and Purwarianti, Ayu and Lestari, Dessi Puji and Azizah, Kurniawati and Tanaya, Dipta and Sakti, Sakriani},
|
15 |
+
title = {Speech Recognition and Meaning Interpretation: Towards Disambiguation of Structurally Ambiguous Spoken Utterances in Indonesian},
|
16 |
+
booktitle = {Proceedings of the EMNLP 2023},
|
17 |
+
year = {2023}
|
18 |
+
}
|
19 |
+
"""
|
20 |
+
|
21 |
+
_DATASETNAME = "struct_amb_ind"
|
22 |
+
|
23 |
+
_DESCRIPTION = """
|
24 |
+
This dataset contains the first Indonesian speech dataset for structurally ambiguous utterances and each of transcription and two disambiguation texts.
|
25 |
+
The structurally ambiguous sentences were adapted from Types 4,5,6, and 10 of Types Of Syntactic Ambiguity in English by [Taha et al., 1983].
|
26 |
+
For each chosen type, 100 structurally ambiguous sentences in Indonesian were made by crowdsourcing.
|
27 |
+
Each Indonesian ambiguous sentence has two possible interpretations, resulting in two disambiguation text outputs for each ambiguous sentence.
|
28 |
+
Each disambiguation text is made up of two sentences. All of the sentences have been checked by linguists.
|
29 |
+
"""
|
30 |
+
|
31 |
+
_HOMEPAGE = "https://github.com/ha3ci-lab/struct_amb_ind"
|
32 |
+
|
33 |
+
_LICENSE = Licenses.UNKNOWN.value
|
34 |
+
|
35 |
+
_LOCAL = True # get the audio data externally from https://drive.google.com/drive/folders/1QeaptstBgwGYO6THGkZHHViExrogCMUj
|
36 |
+
_LANGUAGES = ["ind"]
|
37 |
+
|
38 |
+
_URL_TEMPLATES = {
|
39 |
+
"keys": "https://raw.githubusercontent.com/ha3ci-lab/struct_amb_ind/main/keys/train_dev_test_spk_keys/",
|
40 |
+
"text": "https://raw.githubusercontent.com/ha3ci-lab/struct_amb_ind/main/text/",
|
41 |
+
}
|
42 |
+
|
43 |
+
_URLS = {
|
44 |
+
"split_train": _URL_TEMPLATES["keys"] + "train_spk",
|
45 |
+
"split_dev": _URL_TEMPLATES["keys"] + "dev_spk",
|
46 |
+
"split_test": _URL_TEMPLATES["keys"] + "test_spk",
|
47 |
+
"text_transcript": _URL_TEMPLATES["text"] + "ID_amb_disam_transcript.txt",
|
48 |
+
}
|
49 |
+
|
50 |
+
_SUPPORTED_TASKS = [Tasks.SPEECH_RECOGNITION]
|
51 |
+
|
52 |
+
_SOURCE_VERSION = "1.0.0"
|
53 |
+
|
54 |
+
_SEACROWD_VERSION = "2024.06.20"
|
55 |
+
|
56 |
+
|
57 |
+
class StructAmbInd(datasets.GeneratorBasedBuilder):
|
58 |
+
"""
|
59 |
+
This dataset contains the first Indonesian speech dataset for structurally ambiguous utterances and each of transcription and two disambiguation texts.
|
60 |
+
This dataloader does NOT contain the additional training data for as mentioned in the _HOMEPAGE, as it is already implemented in the dataloader "indspeech_news_lvcsr".
|
61 |
+
"""
|
62 |
+
|
63 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
64 |
+
SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
|
65 |
+
|
66 |
+
BUILDER_CONFIGS = [
|
67 |
+
SEACrowdConfig(
|
68 |
+
name=f"{_DATASETNAME}_source",
|
69 |
+
version=SOURCE_VERSION,
|
70 |
+
description=f"{_DATASETNAME} source schema",
|
71 |
+
schema="source",
|
72 |
+
subset_id=f"{_DATASETNAME}",
|
73 |
+
),
|
74 |
+
SEACrowdConfig(
|
75 |
+
name=f"{_DATASETNAME}_seacrowd_sptext",
|
76 |
+
version=SEACROWD_VERSION,
|
77 |
+
description=f"{_DATASETNAME} SEACrowd schema",
|
78 |
+
schema="seacrowd_sptext",
|
79 |
+
subset_id=f"{_DATASETNAME}",
|
80 |
+
),
|
81 |
+
]
|
82 |
+
|
83 |
+
DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
|
84 |
+
|
85 |
+
def _info(self) -> datasets.DatasetInfo:
|
86 |
+
if self.config.schema == "source":
|
87 |
+
features = datasets.Features(
|
88 |
+
{
|
89 |
+
"id": datasets.Value("string"),
|
90 |
+
"speaker_id": datasets.Value("string"),
|
91 |
+
"path": datasets.Value("string"),
|
92 |
+
"audio": datasets.Audio(sampling_rate=16_000),
|
93 |
+
"amb_transcript": datasets.Value("string"),
|
94 |
+
"disam_text": datasets.Value("string"),
|
95 |
+
}
|
96 |
+
)
|
97 |
+
|
98 |
+
elif self.config.schema == "seacrowd_sptext":
|
99 |
+
features = schemas.speech_text_features
|
100 |
+
|
101 |
+
return datasets.DatasetInfo(
|
102 |
+
description=_DESCRIPTION,
|
103 |
+
features=features,
|
104 |
+
homepage=_HOMEPAGE,
|
105 |
+
license=_LICENSE,
|
106 |
+
citation=_CITATION,
|
107 |
+
)
|
108 |
+
|
109 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
|
110 |
+
# The data_dir configuration is required ONLY for the audio_urls.
|
111 |
+
if self.config.data_dir is None:
|
112 |
+
raise ValueError("This is a local dataset. Please pass the data_dir kwarg to load_dataset.")
|
113 |
+
else:
|
114 |
+
data_dir = self.config.data_dir
|
115 |
+
|
116 |
+
# load the local audio folders
|
117 |
+
audio_urls = [data_dir + "/" + f"{gender}{_id:02}.zip" for gender in ["F", "M"] for _id in range(1, 12, 1)]
|
118 |
+
audio_files_dir = [Path(dl_manager.extract(audio_url)) / audio_url.split("/")[-1][:-4] for audio_url in audio_urls]
|
119 |
+
# load the speaker splits and transcript
|
120 |
+
split_train = Path(dl_manager.download(_URLS["split_train"]))
|
121 |
+
split_dev = Path(dl_manager.download(_URLS["split_dev"]))
|
122 |
+
split_test = Path(dl_manager.download(_URLS["split_test"]))
|
123 |
+
text_transcript = Path(dl_manager.download(_URLS["text_transcript"]))
|
124 |
+
|
125 |
+
return [
|
126 |
+
datasets.SplitGenerator(
|
127 |
+
name=datasets.Split.TRAIN,
|
128 |
+
gen_kwargs={"split": split_train, "transcript": text_transcript, "audio_files_dir": audio_files_dir},
|
129 |
+
),
|
130 |
+
datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"split": split_dev, "transcript": text_transcript, "audio_files_dir": audio_files_dir}),
|
131 |
+
datasets.SplitGenerator(
|
132 |
+
name=datasets.Split.TEST,
|
133 |
+
gen_kwargs={"split": split_test, "transcript": text_transcript, "audio_files_dir": audio_files_dir},
|
134 |
+
),
|
135 |
+
]
|
136 |
+
|
137 |
+
def _generate_examples(self, split: Path, transcript: Path, audio_files_dir: List[Path]) -> Tuple[int, Dict]:
|
138 |
+
speaker_ids = open(split, "r").readlines()
|
139 |
+
speaker_ids = [id.replace("\n", "") for id in speaker_ids]
|
140 |
+
speech_folders = [audio_folder for audio_folder in audio_files_dir if audio_folder.name.split("/")[-1] in speaker_ids]
|
141 |
+
speech_files = list(chain(*[list(map((str(speech_folder) + "/").__add__, os.listdir(speech_folder))) for speech_folder in speech_folders]))
|
142 |
+
|
143 |
+
transcript = open(transcript, "r").readlines()
|
144 |
+
transcript = [sent.replace("\n", "").split("|") for sent in transcript]
|
145 |
+
transcript_dict = {sent[0]: {"amb_transcript": sent[1], "disam_text": sent[2]} for sent in transcript}
|
146 |
+
|
147 |
+
for key, aud_file in enumerate(speech_files):
|
148 |
+
aud_id = aud_file.split("/")[-1][:-4]
|
149 |
+
aud_info = aud_id.split("_")
|
150 |
+
if self.config.schema == "source":
|
151 |
+
row = {
|
152 |
+
"id": aud_id,
|
153 |
+
"speaker_id": aud_info[1],
|
154 |
+
"path": aud_file,
|
155 |
+
"audio": aud_file,
|
156 |
+
"amb_transcript": transcript_dict[aud_id]["amb_transcript"],
|
157 |
+
"disam_text": transcript_dict[aud_id]["disam_text"],
|
158 |
+
}
|
159 |
+
yield key, row
|
160 |
+
elif self.config.schema == "seacrowd_sptext":
|
161 |
+
row = {
|
162 |
+
"id": aud_id,
|
163 |
+
"path": aud_file,
|
164 |
+
"audio": aud_file,
|
165 |
+
"text": transcript_dict[aud_id]["amb_transcript"],
|
166 |
+
"speaker_id": aud_info[1],
|
167 |
+
"metadata": {
|
168 |
+
"speaker_age": None,
|
169 |
+
"speaker_gender": aud_info[1][0],
|
170 |
+
},
|
171 |
+
}
|
172 |
+
yield key, row
|
173 |
+
else:
|
174 |
+
raise NotImplementedError(f"Schema '{self.config.schema}' is not defined.")
|