Transcription voice recognition engine added! Added supported file formats! Mr. Transcription powers up! June / July 2021 Update

Aug. 2, 2021

Transcription voice recognition engine added! Added supported file formats! Mr. Transcription powers up! June / July 2021 Update | AI character transcription service - Mr. Transcription

It is the development team of Mr. Transcription who is working hard on development today.

Transcription was released in May, and minor updates are made daily.

In addition, in June and July, we made a major update to Mr. Transcription.

Many people have noticed the update because the transcription part of the main function has changed a lot.

However, I am also fixing other small details.

How much have you noticed the update?

We will introduce the details of the updates made in June and July.

Major updates in June and July

  1. Added supported file formats
  2. Added voice recognition engine
  3. Added 1 minute trial transcription with 2 voice recognition engines
  4. Addition of dictionary function
  5. Other updates

Added supported file formats

Corresponding file

At the time of the May release, the only file formats supported were very limited.

The files currently available are as follows.

  • 【Image file】
    Supported file formats .jpg .jpeg .png .webp
    * If the characters are facing sideways or downwards, you cannot transcribe them, so please correct the direction.
  • [Audio file]
    Audio time 90 minutes or less Supported file formats .mp3 .wav .wma .mp4 .aifc .flac .aac .aiff .aifc
  • [Video file]
    Audio time 90 minutes or less Supported file formats .mp4 .mov .avi /flv .mkv .webm .wmv .3gp

With the update, it supports various file formats, and you no longer need to convert files yourself before using it.

There are plans to increase the number of compatible files such as PDF files in the future.

Added voice recognition engine


You can now use the voice recognition engine "AmiVoice", which has a reputation for being the most accurate transcription of Japanese voice.

What is AmiVoice? <br /> AmiVoice provided by Advanced Media, Inc. is a voice recognition technology with the largest market share * in Japan. The latest deep learning technology and LSTM (Long Short-Term Memory) technology of recurrent neural network, which is an advanced technology of deep learning, are implemented in the engine. It boasts a high recognition rate.
* Source: ITR "ITR Market View: AI Market 2020" Voice recognition market Share of sales by vendor (2015-2020 forecast)

From the voice recognition pull-down menu

  • Trial (up to 1 minute)
  • AmiVoice
  • Google

And you can choose which engine to use for voice recognition (transcription).

  • Is it transcribed with AmiVoice?
  • Is it transcribed by Google's voice recognition engine?

With more choices, you can feel the wonder of transcription by AI.

Added 1 minute trial transcription with 2 voice recognition engines


With this update, Transcription

  • Google
  • AmiVoice

You can now choose between two types of speech recognition engines.


Which voice recognition engine should be used for transcription ...? Which one can be transcribed better?

Many people should be worried.

Therefore, we have added a function that allows any plan to select "Trial Transcription (1 minute)".

With "Trial (1 minute)", you can transcribe "the first minute" of any length of audio / video files with two engines, AmiVoice and Google.

You can check the accuracy of which voice recognition engine is better. Does the accuracy change so much depending on the voice recognition engine? It is an interesting function that you can realize.

Detailed usage of the trial function is introduced in this article.

[New function] Trial (1 minute) How to use transcription! Transcription with two speech recognition engines at the same time.

Addition of dictionary function


There is a scene where "I want you to hear this word properly" during voice recognition.

A convenient dictionary function has been added in such cases.

The dictionary function allows you to pre-register specific words, such as technical terms, that are usually difficult to transcribe, making it easier for you to transcribe those words properly.

However, if you use it too much, the transcription will be pulled by the registered dictionary word, and even another word may be mistakenly recognized as a dictionary registered word and you may not be able to transcribe it correctly.

Other notes and detailed usage are introduced here.

[New function] How to register the dictionary function of Mr. Transcription. Convenient usage and precautions

Other updates

In addition, we have added internal processing and made minor corrections such as corrections to the TOP page.

The development team will continue to develop so that Mr. Transcription can be used even more comfortably.

We are pleased that it is the easiest AI transcription service for you.

Please continue to use it a lot. Please provide some support.

■ AI transcription service "Mr. Transscription"

"Mr. Transcription" is an online transcription tool that can be used from zero initial cost and 1,000 yen per month (* free version available).

  • Supports more than 20 file formats such as audio, video, and images
  • Can be used from both PC and smartphone
  • Supports technical terms such as medical care, IT, and long-term care
  • Supports creation of subtitle files and speaker separation
  • Supports transcription in approximately 100 languages ​​including English, Chinese, Japanese, Korean, German, French, Italian, etc.

To use it, just upload the audio file from the site. Transcription text is available in seconds to tens of minutes.
You can use it for free if you transcribe it for up to 10 minutes, so please try it once.

It is "Mr. Transcription" who can easily transcribe from audio, video, and images. Transcription allows you to transcribe for up to 10 minutes for free. You can copy, download, search, delete, etc. the transcribed text. You can also create subtitle files, which is ideal for transcription of interview videos.
Related article