How to get Google to transcribe videos and provide subtitles?

Question

Google Voice Input and Google Translate seems to have the ability to recognize speech voice from recorded voice. It doesn't work on recorded video. I like to know if there is a way to enable such feature so I can understand some of the recorded video documents. If not, is there some other service that can do audio input from playing a recording?

Answer 1

Upload the video to Youtube, keep it marked private if you don't want the world to see it, and let google have time with it, meaning it may take a few moments or a few hours, and Youtube (google) will create Auto Transcripts.

If its audio only, look for a Photo Video Maker (free, online, etc) and add your audio track to a single static picture or multiple pictures.

Caveat, nothing will help bad audio without professional tweaks.

Answer 2

Your best bet is to use autosub. Essentially it analyses your video and provides you with subtitles using YouTube's method.

1st step: Install chocolatey as admin. Open a prompt command window and type:

@"%SystemRoot%\System32\WindowsPowerShell\v1.0\powershell.exe" -NoProfile -InputFormat None -ExecutionPolicy Bypass -Command "iex ((New-Object System.Net.WebClient).DownloadString('https://chocolatey.org/install.ps1'))" && SET "PATH=%PATH%;%ALLUSERSPROFILE%\chocolatey\bin"

After installation is done, go to a prompt command window and type:

choco install pip python27 ffmpeg -y

2nd step: Install autosub type refreshenv

now type pip install autosub

Test using:

C:\Python27\python.exe C:\Python27\scripts\autosub_app.py -S en -D en YourVideo.mp4 

Answer 3

The problem is that Google translate is listening to the microphone. That means you either broadcast with speakers and have your computer microphone listen in. This introduces all kinds of noise and physical problems. So what you need is for the computer to use the line input and listen in to the audio output (ie whatever you are playing on the computer). The solution is a free software called VoiceMeeter (google search for the free download). Once you install it, you can open up your Speaker symbol on your bottom tool bar (choose playback devices, Recording) and you will notice on the microphone input there are 2 entries. One is your microphone and the other is your VoiceMeeter Output. Choose VoiceMeeter Output . When you switch to playback, you will have headphones, Digital Audio and VoiceMeeter Input. Choose VoiceMeeter Input.

Therefore You when you play a video, the audio is being fed into VoiceMeeter Input and VoiceMeeter is feeding it into the Microphone. Now Activate the microphone in Google Translate, choose the language and watch it translate into the right pane in english. If you want to save the text you have to highlight it and capture it into your Word or else its lost. Adjust the volume. Sometimes its too loud and translate will not work.

I was able to translate some japanese programs this way and understand what they were saying. The translation is not perfect but better than anything else.

Note that you will not hear anything when you do this as the audio system is getting no input. Opening up VoiceMeeter you have a full sound processor so you can increase the base, trebble etc. I havent played around with it enough but you might be able to add a 2nd line to the speakers so that you can hear it while it is translating.

Hope this helps someone.