EdwardHayashi-2023
commited on
Commit
•
a263eb7
1
Parent(s):
fd950f4
Update MELD-Audio.py
Browse files- MELD-Audio.py +15 -7
MELD-Audio.py
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
#!/usr/bin/env python3
|
2 |
# -*- coding: utf-8 -*-
|
3 |
"""
|
4 |
-
Created on
|
5 |
|
6 |
@author: lin.kinwahedward
|
7 |
"""
|
@@ -42,6 +42,7 @@ _HOMEPAGE = "https://affective-meld.github.io/"
|
|
42 |
|
43 |
_LICENSE = "CC BY 4.0"
|
44 |
|
|
|
45 |
_DATA_URL = "https://drive.google.com/uc?export=download&id=1TPr9v5Vz1qQuxPWcr8RedfuQvLyuG1lm"
|
46 |
|
47 |
#------------------------------------------------------------------------------
|
@@ -72,16 +73,14 @@ class MELD_Audio(datasets.GeneratorBasedBuilder):
|
|
72 |
#--------------------------------------------------------------------------
|
73 |
'''
|
74 |
Define the "column header" (feature) of a datum.
|
75 |
-
|
76 |
-
1)
|
77 |
-
2)
|
78 |
-
3) emotion label
|
79 |
'''
|
80 |
def _info(self):
|
81 |
|
82 |
features = datasets.Features(
|
83 |
{
|
84 |
-
"path": datasets.Value("string"),
|
85 |
"audio": datasets.Audio(sampling_rate = 16000),
|
86 |
"label": datasets.ClassLabel(
|
87 |
names = [
|
@@ -105,9 +104,17 @@ class MELD_Audio(datasets.GeneratorBasedBuilder):
|
|
105 |
)
|
106 |
#--------------------------------------------------------------------------
|
107 |
def _split_generators(self, dl_manager):
|
|
|
|
|
|
|
|
|
|
|
|
|
108 |
|
|
|
109 |
dataset_path = dl_manager.download_and_extract(self.config.data_url)
|
110 |
|
|
|
111 |
return [
|
112 |
datasets.SplitGenerator(
|
113 |
name = datasets.Split.TRAIN,
|
@@ -132,6 +139,8 @@ class MELD_Audio(datasets.GeneratorBasedBuilder):
|
|
132 |
def _generate_examples(self, audio_path, csv_path):
|
133 |
'''
|
134 |
Get the audio file and set the corresponding labels
|
|
|
|
|
135 |
'''
|
136 |
key = 0
|
137 |
with open(csv_path, encoding = "utf-8") as csv_file:
|
@@ -141,7 +150,6 @@ class MELD_Audio(datasets.GeneratorBasedBuilder):
|
|
141 |
_, _, _, emotion, _, dialogue_id, utterance_id, _, _, _, _ = row
|
142 |
filename = "dia" + dialogue_id + "_utt" + utterance_id + ".mp3"
|
143 |
yield key, {
|
144 |
-
"path": audio_path + filename,
|
145 |
# huggingface dataset's will use soundfile to read the audio file
|
146 |
"audio": audio_path + filename,
|
147 |
"label": emotion,
|
|
|
1 |
#!/usr/bin/env python3
|
2 |
# -*- coding: utf-8 -*-
|
3 |
"""
|
4 |
+
Created on Mon Apr 24 10:45:46 2023
|
5 |
|
6 |
@author: lin.kinwahedward
|
7 |
"""
|
|
|
42 |
|
43 |
_LICENSE = "CC BY 4.0"
|
44 |
|
45 |
+
# The actual place where the data is stored!
|
46 |
_DATA_URL = "https://drive.google.com/uc?export=download&id=1TPr9v5Vz1qQuxPWcr8RedfuQvLyuG1lm"
|
47 |
|
48 |
#------------------------------------------------------------------------------
|
|
|
73 |
#--------------------------------------------------------------------------
|
74 |
'''
|
75 |
Define the "column header" (feature) of a datum.
|
76 |
+
2 Features:
|
77 |
+
1) audio samples
|
78 |
+
2) emotion label
|
|
|
79 |
'''
|
80 |
def _info(self):
|
81 |
|
82 |
features = datasets.Features(
|
83 |
{
|
|
|
84 |
"audio": datasets.Audio(sampling_rate = 16000),
|
85 |
"label": datasets.ClassLabel(
|
86 |
names = [
|
|
|
104 |
)
|
105 |
#--------------------------------------------------------------------------
|
106 |
def _split_generators(self, dl_manager):
|
107 |
+
'''
|
108 |
+
Split the dataset into datasets.Split.{"TRAIN", "VALIDATION", "TEST", "ALL"}
|
109 |
+
|
110 |
+
The dataset can be further modified, please see below link for details.
|
111 |
+
https://huggingface.co/docs/datasets/process
|
112 |
+
'''
|
113 |
|
114 |
+
# Get the dataset and store at the machine where this script is executed!
|
115 |
dataset_path = dl_manager.download_and_extract(self.config.data_url)
|
116 |
|
117 |
+
# "audio_path" and "csv_path" would be the parameters passed to def _generate_examples()
|
118 |
return [
|
119 |
datasets.SplitGenerator(
|
120 |
name = datasets.Split.TRAIN,
|
|
|
139 |
def _generate_examples(self, audio_path, csv_path):
|
140 |
'''
|
141 |
Get the audio file and set the corresponding labels
|
142 |
+
|
143 |
+
Must execute till yield, otherwise, error will occur!
|
144 |
'''
|
145 |
key = 0
|
146 |
with open(csv_path, encoding = "utf-8") as csv_file:
|
|
|
150 |
_, _, _, emotion, _, dialogue_id, utterance_id, _, _, _, _ = row
|
151 |
filename = "dia" + dialogue_id + "_utt" + utterance_id + ".mp3"
|
152 |
yield key, {
|
|
|
153 |
# huggingface dataset's will use soundfile to read the audio file
|
154 |
"audio": audio_path + filename,
|
155 |
"label": emotion,
|