chrisjay commited on
Commit
37c6e3d
β€’
1 Parent(s): 75fd365

worked on motivation

Browse files
Files changed (1) hide show
  1. app.py +5 -2
app.py CHANGED
@@ -246,7 +246,10 @@ This is a platform to contribute to your African language by recording your voic
246
  markdown="""
247
  # 🌍 African Digits Recording Sprint
248
  Existing speech recognition systems do not support ANY African languages, excluding African speakers from voice-enabled devices. Our voice is our identity!
249
- And so the goal of the project is to make a simple digits dataset for African languages through crowd-sourcing. You can easily teach a model to recognise numbers in your language using this dataset.
 
 
 
250
 
251
  """
252
  record_markdown = """
@@ -367,7 +370,7 @@ with block:
367
  #listen = gr.Button("Listen")
368
  listen_tab.select(show_records,inputs=[],outputs=[display_html,plot]+PLOTS_FOR_GRADIO)
369
 
370
- with gr.TabItem('Hear') as listen_tab:
371
  # Dropdown to choose a language from any of the 6
372
  # When you choose, it will load the correponding model
373
  # And then one can record a voice and get the model prediction
 
246
  markdown="""
247
  # 🌍 African Digits Recording Sprint
248
  Existing speech recognition systems do not support ANY African languages, excluding African speakers from voice-enabled devices. Our voice is our identity!
249
+
250
+ The purpose of this project is to show the effectiveness of community-based crowd-sourcing dataset curation in the development of technologies for African languages.
251
+
252
+ We start with a simple digits dataset for African languages through crowd-sourcing. You can easily teach a model to recognise numbers in your language using this dataset.
253
 
254
  """
255
  record_markdown = """
 
370
  #listen = gr.Button("Listen")
371
  listen_tab.select(show_records,inputs=[],outputs=[display_html,plot]+PLOTS_FOR_GRADIO)
372
 
373
+ with gr.TabItem('Model') as listen_tab:
374
  # Dropdown to choose a language from any of the 6
375
  # When you choose, it will load the correponding model
376
  # And then one can record a voice and get the model prediction