iFlytek

iFlytek PC versionIt is a voice note tool with an intuitive interface, outstanding functions, and easy operation. The PC version of iFlytek Yuji can help users convert speech into text and record it, and supports multi-end synchronization. Documents can also be shared, and it supports intelligent proofreading while listening. iFlytek Yuji brings users a different experience and Feel, it can be said that it is a must-have artifact for voice notes. Friends who like iFlytek, please come to Huajun Software Park to download and experience it!

Introduction to the functions of iFlytek Voice Notes

​ ​ 1. Voice input and efficient writing

The world’s cutting-edge voice technology has a recognition rate of up to 98%.

​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​

​​ 120 minutes of continuous voice input, your thoughts will flow uninterrupted.

​ ​ 2. Recording transcription and intelligent proofreading

The recording is converted into text in real time, and you can correct it while listening without missing a word.

​​A must-have tool for reporter interviews, meeting minutes, and class notes.

​​ 3. Multi-terminal login and cloud synchronization

​​Supports multi-terminal login such as iOS, Android, and Web.

​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​?

4. Collection assistant voice reading

​ ​ Collect articles, pictures, links and other content with one click.

                      Read aloud all kinds of reading materials, free your eyes, and there are a variety of pronunciations to choose from.

iFlytek Voicemail installation steps

​​ 1. Download this software from Huajun Software Park, unzip it, and double-click the exe file

2. Set the software installation location and click Install

3. The software is being installed, we are waiting patiently

4. Software installation successful

iFlytek frequently asked questions

​ ​ 1. Questions about frequency limit

The service volume in the development and integration phase is 500 times per day. Developers need to submit their application for online review. After passing the application, they can get more daily service volume.

​ ​ 2. Time range for application online review

The application online review process takes 4 working days (working days do not include Saturdays, Sundays and national statutory holidays).

3. Sampling rate and sampling accuracy of uploaded audio

Audio in PCM or WAV format with a sampling rate of 16KHZ or 8KHZ, mono, and a sampling accuracy of 16bit.

​ 4. Chinese and English mixed recognition

It is not supported yet.

5. Dictation VS Recognition

Dictation: Convert speech directly into corresponding text.

Recognition: First upload the command word or abnf grammar file. If a command word is uploaded, only the command word will be recognized and the recognition result will be returned. If there is no match in the command word, an error code with no matching result will be returned; if a grammar file is uploaded, it will be returned in the format of the grammar file. Corresponding matching information, if there is no matching result, an error code with no matching result will be returned.

​ 6. Using voice dictation, the maximum time of a conversation

The maximum duration of a session is 60 seconds.

​ 7. The difference between contacts, user vocabulary, command words, abnf grammar files, and bnf grammar files

​​​​Contacts and user vocabulary are used during dictation, and command words and abnf grammar files are used during recognition.

When using contacts, the program will read the user's contact information and give priority to identifying the contacts when dictating; use the user vocabulary list, which is customized by the developer and can involve various side or professional terms. Upload the vocabulary list Finally, the words in the vocabulary list will be recognized first to improve the personalized recognition experience.

When recognizing command words, you need to upload the command words. During recognition, the recognition results will only match the command words, and there will be a confidence level in the recognition results; when using abnf grammar files, you need to upload the grammar files, and the grammar files can be written by yourself. During recognition, the result will be returned only if the result matches the grammar file, otherwise an error of no matching result will be returned; when using the offline command word recognition of Voice+, you need to use bnf grammar to write the grammar file locally, and then you can build it locally. Local grammar file recognized.

​ 8. Log file generation configuration

When "output=1" is configured in the msc.cfg file, the running log will be output to msc.log, lvl=-1, log=file path/msc.log (for example./home/../msc.log); "Output=0" does not generate logs. The smaller the Level value, the more detailed the log will be. Level=-1 is the most detailed.

​​ 9.Does the data need to have a wav header? Can the mic data be sent directly to the server?

The wav file consists of a pcm file header and pcm data. The pcm file header is used to identify the pcm data sampling rate, quantization bit number, file length and other information of the wav file.

What is obtained from mic recording is PCM data, which can be directly uploaded to the voice cloud for recognition without adding a PCM file header.

It should be noted that the pcm data sampling rate and quantization bit number of mic recording need to be consistent with the parameters set in SessionBegin.

iFlytek update log

​​ 1. Added two new long picture sharing VIP backgrounds;

​ ​ 2. Optimize the time machine function and restore the version more smoothly;

3. Fix other known bugs