Spaces:
Build error
Build error
updated article.
Browse files- app.py +6 -4
- article.py +10 -1
app.py
CHANGED
@@ -55,6 +55,8 @@ def save_record(language,text,record,number,age,gender,accent,number_history,cur
|
|
55 |
speaker_metadata['gender'] = gender if gender!=GENDER[0] else ''
|
56 |
speaker_metadata['age'] = age if age !='' else ''
|
57 |
speaker_metadata['accent'] = accent if accent!='' else ''
|
|
|
|
|
58 |
if not done_recording:
|
59 |
if language!=None and language!='Choose language' and record is not None and number is not None:
|
60 |
language = language.lower()
|
@@ -131,7 +133,7 @@ def save_record(language,text,record,number,age,gender,accent,number_history,cur
|
|
131 |
output_string = "<div class='finished'>"+output+"</div>"
|
132 |
next_number = 0 # the default number
|
133 |
next_number_image = f'number/best.gif'
|
134 |
-
return output_string,next_number_image,number_history,next_number,done_recording
|
135 |
|
136 |
def display_records():
|
137 |
repo.git_pull()
|
@@ -182,15 +184,15 @@ This is a platform to contribute to your African language by recording your voic
|
|
182 |
markdown="""
|
183 |
# 🌍 African Digits Recording Sprint
|
184 |
|
185 |
-
>
|
186 |
|
187 |
1. Choose your African language
|
188 |
2. Fill in the speaker metadata (age, gender, accent). This is optional but important to build better speech models.
|
189 |
3. You will see the image of a number __(this is the number you will record)__.
|
190 |
-
4. Fill in the word of that
|
191 |
5. Click record and say the number in your African language.
|
192 |
6. Click ‘Submit’. It will save your record and go to the next number.
|
193 |
-
7. Repeat
|
194 |
8. Leave a ❤ in the Space, if you found it fun.
|
195 |
"""
|
196 |
|
|
|
55 |
speaker_metadata['gender'] = gender if gender!=GENDER[0] else ''
|
56 |
speaker_metadata['age'] = age if age !='' else ''
|
57 |
speaker_metadata['accent'] = accent if accent!='' else ''
|
58 |
+
|
59 |
+
|
60 |
if not done_recording:
|
61 |
if language!=None and language!='Choose language' and record is not None and number is not None:
|
62 |
language = language.lower()
|
|
|
133 |
output_string = "<div class='finished'>"+output+"</div>"
|
134 |
next_number = 0 # the default number
|
135 |
next_number_image = f'number/best.gif'
|
136 |
+
return output_string,next_number_image,number_history,next_number,done_recording,default_record
|
137 |
|
138 |
def display_records():
|
139 |
repo.git_pull()
|
|
|
184 |
markdown="""
|
185 |
# 🌍 African Digits Recording Sprint
|
186 |
|
187 |
+
> Record numbers 0-9 in your African language.
|
188 |
|
189 |
1. Choose your African language
|
190 |
2. Fill in the speaker metadata (age, gender, accent). This is optional but important to build better speech models.
|
191 |
3. You will see the image of a number __(this is the number you will record)__.
|
192 |
+
4. Fill in the word of that number (optional)
|
193 |
5. Click record and say the number in your African language.
|
194 |
6. Click ‘Submit’. It will save your record and go to the next number.
|
195 |
+
7. Repeat 4-7
|
196 |
8. Leave a ❤ in the Space, if you found it fun.
|
197 |
"""
|
198 |
|
article.py
CHANGED
@@ -7,12 +7,21 @@ Existing speech recognition services are not available in many African languages
|
|
7 |
|
8 |
This dataset will boost speech technologies (like speech-to-text, text-to-speech, speech translation, and modeling) for African languages, which hitherto had little or no public dataset.
|
9 |
|
10 |
-
**Note:** This is a continuous effort.
|
11 |
|
12 |
**Benefits of such a dataset**
|
13 |
- Useful dataset to introduce people to audio-related Machine Learning. It can be used as a simple training and/or evaluation dataset for speech processing tasks.
|
14 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
15 |
**Contact**
|
|
|
16 |
In case of questions, issues or anything contact Chris Emezue at:
|
17 |
- chris@huggingface.co
|
18 |
|
|
|
7 |
|
8 |
This dataset will boost speech technologies (like speech-to-text, text-to-speech, speech translation, and modeling) for African languages, which hitherto had little or no public dataset.
|
9 |
|
10 |
+
**Note:** This is a continuous effort. This sprint is just to kick-start the event.
|
11 |
|
12 |
**Benefits of such a dataset**
|
13 |
- Useful dataset to introduce people to audio-related Machine Learning. It can be used as a simple training and/or evaluation dataset for speech processing tasks.
|
14 |
|
15 |
+
**About the dataset**
|
16 |
+
|
17 |
+
The data (metadat,text, and audio recording) are uploaded to [a public Hugging Face dataset](https://huggingface.co/datasets/chrisjay/crowd-speech-africa).
|
18 |
+
|
19 |
+
We do not collect your name, address or other sensitive information.
|
20 |
+
|
21 |
+
If for some reason you want to remove your entry, please reach out by email.
|
22 |
+
|
23 |
**Contact**
|
24 |
+
|
25 |
In case of questions, issues or anything contact Chris Emezue at:
|
26 |
- chris@huggingface.co
|
27 |
|