Convert voice to text automatically! Introducing VOITER and VOITER mini

What is the appeal of the AI ​​writing recorder recommended for entrepreneurs who want to work efficiently?

The VOITER and VOITER mini, AI writing recorders that automatically convert voice to text, have been released and are becoming a hot topic among entrepreneurs and writers who want to work efficiently. VOITER, which brings together advanced speech recognition technology, and VOITER mini, which supports text separation for each speaker. We will introduce the features of the two models, including interviews with those who actually use them.

*In the "Founding Techo" where I am writing this article, more detailed information is also explained in the thick "Founding Techo / Printed Edition". You can get it for free, so please order it.

Contents of this article

  • Four features of "VOITER" and "VOITER mini"
  • Billing plans for "VOITER" and "VOITER mini"
  • How comfortable is the new product "VOITER"?

    There are many scenes where businessmen need transcription, such as creating minutes of important meetings, recording lectures and training. Among them, writers who make a living doing interviews spend a lot of time transcribing several times a week. We asked Mr. Takanashi, a writer who tried using VOITER, about his impressions.

    I was surprised at how good the AI ​​automatically corrected

    ──Please tell us what made you decide to use VOITER.

    Takanashi: Due to my work, I have tried many apps and products that can automatically transcribe text, but I haven't been able to find one that is highly accurate. At that time, I heard that a senior writer was praising VOITER, which I found on crowdfunding Makuake, so I watched the promotional video for the product. It looked really good, and I thought it might be practical, so I borrowed it and used it.

    The first thing that surprised me when I received it was the small size of the VOITER itself. It's a compact size that fits comfortably in the palm of your hand, and it feels like an iPhone a long time ago. I saw the VOITER mini at a consumer electronics store, but the mini is even smaller and looks like an eraser. The design is inspired by a cassette tape, so it's cute and retro. Moreover, it comes with 4 types of stickers, so you can change it to your liking. Both VOITER and VOITER mini are small enough to be easily carried around, which is also nice.

    ──How was it when you actually used it?

    Takanashi: I've used it many times in interviews, but the AI's automatic correction is really excellent. While looking at the transcribed text data, I listen to the source material again to confirm the content, but about 70% can be used as is. Of course, the speaker's habit of "uh" or "maa" will be included as it is, but there is no problem if you search for unnecessary words in Word and delete them all at once. I always felt that the time spent transcribing was wasted, so I was happy and happy to think that I was freed from the stress of many years. Ever since I learned about VOITER, I have been recommending it to fellow writers and editors whenever I meet them.

    ──iFLYTEK, which developed the product, is China's top manufacturer of voice AI, and ranked 6th in the MIT (Massachusetts Institute of Technology, USA) Technology Review's "50 Smartest Companies in the World" (2017). Because it is. You are a notable company that has been introduced as the top destructive power in China next to Amazon and Google, but do you feel that the quality is satisfactory?

    Takanashi: I learned about that line before I even started using it, and it piqued my interest (laughs). Of course, the quality is acceptable.

    Automatically convert voice to text! VOITER and VOITER mini debut

    You can use it as if you were using a smartphone

    ──How is the usability?

    Takanashi: It's a simple specification, so you can use it just like operating a smartphone without looking at the manual. First, select the most suitable recording mode from the six recording modes of "coverage", "meeting", "lecture", "memo", "music", and "standard". Transcription feature is not available). It can be operated intuitively and there is no waste.

    ──I heard that the transcription will be done at the same time as the recording.

    Takanashi: As long as you are connected to Wi-Fi, you can transcribe the recording at the same time. Then, if you turn on "Display transcription in real time", the text will be displayed instantly on the VOITER screen. However, depending on the interview environment, there may be times when you cannot connect to the internet, so VOITER may record only the audio on the spot and transcribe it later.

    ──Could you tell us more about that?

    Takanashi: If you want to transcribe later, after recording, connect to the internet and the transcription will start automatically. If the server is not congested, it seems that text conversion will be done within about 20 minutes by pressing the "accelerate" button that gives priority to processing. Whether it's done in real time or transcribed later, it seems that the transcription process is being done again in order to improve the accuracy, and the data will be even more complete here. AI thinks according to the context before and after, so it can properly convert homonyms such as "intention" and "transition" into kanji.

    ──That is certainly excellent. Can text files saved on the terminal be easily imported into the computer?

    Takanashi: Yes. There are three ways to import data. One is to copy the data by connecting the VOITER main unit and a computer with a USB cable. If you are using a Mac, you need to install an Android file transfer app in advance, so I haven't tried this method as a Mac user.

    The second method is to share files by email. Simply select a file, click the "Email Share" button, select the type of file you want to share (text only/audio only/both audio and text) and enter the destination address. Click on the link in the email you receive to instantly download the file.

    The third method is to access the dedicated cloud page called "RECORDER STATION" and download the data. Since you access your own page, it's convenient to see the list of files you've uploaded so far.

    I use this method, but it takes a little time to download both voice and text, so I only download text data to my computer. If I need to listen to the sound source again, I listen directly from the VOITER itself. VOITER doesn't have an earphone jack, but I can use Bluetooth, so I connect AirPods and use it.

    VOITER has 16GB storage, mini has 32GB storage

    ──What about the sound quality of the recorded voice?

    Takanashi: It's easy to hear. Even though it's so compact, VOITER has 8 microphones, right?

    ──That's right. Two directional mics with up to 10m speaking distance, and six omnidirectional mics with 360-degree sound pickup are amazing. What's more, the recording quality is automatically identified in real time, and the sound collection role of each microphone is changed to reduce variations in recording quality. Before purchasing, did you care about the capacity of the device itself?

    Takanashi: It didn't bother me at all. VOITER can go up to 16GB, and VOITER mini seems to be OK up to 32GB, so no one has a problem, right? I don't think there will be a shortage of data, as the data will be downloaded to the computer immediately after the interview. The cloud page for uploading data can also be saved up to 10GB per account, so I use it without worrying about the capacity.

    Can VOITER automatically add subtitles to videos?

    ──Did you know that VOITER not only transcribes the recorded audio, but also automatically adds subtitles to the video?

    Takanashi: No. You can even do that! A friend of mine has uploaded a lecture video on YouTube, and I was just talking about how it would be easier to convey if a caption was included. Shall I tell you quickly?

    ──How about the price range?

    Takanashi: If you can shorten the work time so much, this price is about cheap. VOITER mini is a little cheaper, but the maximum sound collection distance is 5m, so I tried using VOITER, which can reliably convert a wider range of sound into text (VOITER has a maximum sound collection distance of 10m). But if you think about it, 5m is enough for a purpose like mine. Because in an interview there are at least two speakers.

    ──It's hard to decide which one to choose. Finally, what kind of people would you recommend the VOITER or VOITER mini to?

    Takanashi: I would like to recommend it not only to writers and editors in the same industry, but also to those who need to take the minutes of meetings. The time spent transcribing audio into text yields nothing, it's just a waste of time working. If you watch the product's promotional video, you'll be surprised at how accurate it is, and I'm sure you'll want one!

    Four features of "VOITER" and "VOITER mini"

    From here, let's look at the four features of VOITER and VOITER mini.

    Feature 1: Equipped with No. 1 award-winning technology in an international contest that competes for speech recognition performance in noisy environments

    The international contest "CHiME" will be held three times from 2016 to 2020 (held every other year) Equipped with iFLYTEK's high-level speech recognition technology, which has won consecutive No. 1 awards, and advanced noise processing technology that supports it. The quality is demonstrated even in noisy places where a large number of people gather.

    Feature 2: Maximum sound collection distance of 10m (VOITER mini: 5m), 360-degree sound collection and text

    Maximum sound collection distance: 10m (VOITER mini: 5m), 360-degree collection Since it can be used for sound, there is no need to worry about conferences, lectures, etc.

    Feature 3: VOITER supports video subtitles and automatic subtitle generation

    VOITER is equipped with an 8-megapixel camera that automatically generates subtitles while shooting when connected to a network. You can put it on. Subtitles can be generated later after shooting offline. *Only vertical videos are supported. VOITER mini is not supported

    Feature 4: VOITER mini supports text separation for each speaker

    VOITER mini can separate the recorded voice for each speaker and convert it to text, making it easier to record meeting minutes and lecture notes. You can create *VOITER is not supported as of October 2021

    Price plans for "VOITER" and "VOITER mini"

    Transcribing will incur a usage fee depending on the time. In the first year, the following advantageous plans are included free of charge.

    From the founding notebook, we will deliver detailed materials about "VOITER" and "VOITER mini" free of charge. Please use the button below to request information.

    Request more information

    (supervisor: iFLYTEK JAPAN AI SOLUTIONS Co., Ltd.) (edit: founding notebook editorial department)

    Articles that everyone is reading in this category No. 1 convenient apps/servicesYou can quickly find out which subsidies/subsidies are suitable for your company! Let's introduce an automatic matching tool 2nd place Useful apps/services Need a fixed-line phone for a corporation? We will recommend and compare! 3rd place Convenient application service How to make your own business card easily for free & 12 recommended templates / creation tools