Audio-Enabled Question Answering

Convert between audio signals and text transcriptions

Start by downloading the text of presidential inaugural speeches:

In[1]:=

ro=ResourceObject["Presidential Inaugural Addresses"]

Out[1]=

ResourceObject

Name: Presidential Inaugural Addresses »
Type: DataResource
Description: Presidential inaugural addresses through 2017



Select a part of one of the speeches to use as the context for question answering:

In[2]:=

context=StringTakeSelectFirstNormal[ResourceData[ro]],#Name

Lyndon B. Johnson

PERSON

&["Text"],{309,661}

Out[2]=

For every generation there is a destiny. For some, history decides. For this generation the choice must be our own.Even now, a rocket moves toward Mars. It reminds us that the world will not be the same for our children, or even for ourselves in a short span of years. The next man to stand here will look out on a scene that is different from our own.

Now record a question using

AudioCapture

In[3]:=

recordedQuestion=AudioCapture[]

Out[3]=

00:00

00:04

Data in Notebook

Transcribe the recorded question with

SpeechRecognize

In[4]:=

question=SpeechRecognize[recordedQuestion]

Out[4]=

why is the rocket significant

Use

FindTextualAnswer

to identify the most probable answer to the question:

In[5]:=

answer=FindTextualAnswer[context,question,1,"HighlightedParagraph"]

Out[5]=

For every generation there is a destiny. For some, history decides. For this generation the choice must be our own.Even now, a rocket moves toward Mars.

It reminds us that the world will not be the same for our children, or even for ourselves in a short span of years

. The next man to stand here will look out on a scene that is different from our own.

Use

SpeechSynthesize

to convert the result into an

Audio

object. Choose two different voices from

$VoiceStyles

to represent the surrounding context and answer, respectively:

In[6]:=

speakers=RandomSample[$VoiceStyles,2];

Define a function to synthesize speech using a specific voice for the given input:

In[7]:=

synthesize[s_String]:=SpeechSynthesize[s,speakers〚1〛];synthesize[Highlighted[s_String]]:=SpeechSynthesize[s,speakers〚2〛];

Synthesize and combine parts of the speech:

In[8]:=

AudioJoin[synthesize/@First[First[answer]]]

Out[8]=

00:00

00:20

Data in File[

...7e1-4ffeefe6e963.w

]

Publisher Information

Contributed by: Wolfram Staff

Wolfram Language Example Repository

Audio-Enabled Question Answering

See Also

Publisher Information

Audio-Enabled Question Answering

See Also

Related Symbols

Publisher Information