Let's improve customer satisfaction by using case examples of reading aloud software
Jan. 17, 2021
Text-to-speech software is far more familiar than we think.
Speaking of recent famous ones
- Amazon Alexa,
- Google Home
There is an AI assistant in these units.
When you ask a question, the AI assistant will answer you.
But that doesn't mean there are humans inside responding.
Here are some examples of how to use text-to-speech software.
What is text-to-speech software?
Text-to-speech software is a software that converts documents such as text and characters into speech and reads them aloud .
In addition to reading in Japanese, there are also multilingual versions such as English, Chinese, German, Spanish, and Italian.
You can change the speed of the sound freely, and some can be downloaded as an audio file such as .mp3.
About text-to-speech composition
The voice of the AI assistant is composed by text reading software.
For example, let's say an AI assistant replies, "Today's weather is sunny."
In the machine, the letters that reply "kyo, u, no, te, n, ki, wa, ha, re, de, su" are instantly assembled.
Then, by calling out each character "sound" and uttering it, it is possible to output and reply that "Today's weather is sunny" as a voice.
However, since it is too unnatural to speak one character at a time, inside the software
- Voice processing,
- Voice editing
And so on, the decorations are made so that the person is as close as possible to the word actually pronounced.
Specifically, intonation and joints.
Depending on the text-reading software, the intonation and the decoration of the joint may differ.
So, depending on the company that provides the text-to-speech software, the way people perceive "easy to hear" and "easy to use" sound is different.
Scenes that requires text-to-speech software
Text-to-speech software is generally for
Elderly people who have difficulty reading small letters
There were many people who were thought to be needed.
But in reality it is needed in more situations than we think.
If you look at the case studies, you can clearly see the necessary scenes.
Example of using text-to-speech software
- As an automatic voice such as answering machine service
- As a narration for videos such as Youtube
- As a means for the blind
- As an emergency announcement
- As a radio role
1. As an automatic voice for answering service
The format is fixed
- Answering machine,
- Calling guidance service with automatic voice,
- Video audio for in-house training,
For example, text-to-speech software is used to read out fixed-form sentences.
By adding tonality, it is possible to bring people closer to what they are talking to, so the sound quality of telephones and other devices is at a level that is not much different from human voice.
2. As a narration for videos such as Youtube
With the popularity of Youtube, there have been many cases of using text-to-speech software as narration for Youtube.
Even in the "slow" video, the narration uses text-to-speech software.
In some TV programs, text-to-speech software is also used for narration.
3. As a means for the blind
It is difficult for people who are blind to read aloud, even if there is text in it, without support.
Therefore, the text-to-speech software has been used for a long time.
My acquaintance was always listening at 4x speed. Wow...
4. As an emergency announcement
Did you know that text-to-speech software is also used in J-ALERT (national instant warning system)?
It is very important to broadcast in the town to call residents to evacuate in the event of an emergency disaster.
However, in order to broadcast, it was necessary to have people on the spot.
Now it's possible to broadcast with text-to-speech software, to keep staff safe and to do other work.
5. As a radio role
Have you ever heard about "doing while listening"?
While doing housework, commuting, and doing other things, you can get different information from your ears.
The service specialized in "while listening" is "Arukiki" from Asahi Shimbun.
This service is perfect for busy people who can get important news about the day in about 5 minutes.
This news is read by text-to-speech software, not by humans.
Text-to-speech software is familiar
The text-to-speech software is actually more familiar than everyone thinks.
"I'm not good at mechanical noise"
"Difficult to hear"
There are many people who think that, but recently, the tonality has been improved and it is possible to talk more like a person.
Also, reading aloud is more convenient than you think.
For example, in this blog
- Reading silently,
- Listening only by voice,
- While listening to the voice, follow the sentence with your eyes,
Then, it is completely different in readability and ease of coming into your head.
There are also research results that indicate that the more humans use the five senses, the more likely they are to be impressed.
Reading aloud is very convenient.
For example, by adding a reading function to blogs as well, the added value is that "this blog can be read aloud." Added value leads to improved customer satisfaction.
Convenient things are easy for people to accept and easy to spread.
In the future, the "sentence reading" service will spread widely.
"Ondoku" is a Text-to-Speech service that anyone can use for free without installation. You can also use an extension ( Ondoku3-ChatGPT ) that allows you to talk with ChatGPT for free.