Getting unique content from video blogs

The subject of shorthand is not new, so we will use it to get a unique text.
I did not find a specialized software for this task as a single program and decided to use several programs for implementation:

1) RealSpeaker PRO 1.5

2) Virtual Audio Cable 4.10 Full
')
3) SplitCam

4) Text editor (Notepad, Word, etc.)


A little about the software.

The main program we will have is RealSpeakerPRO 1.5 which will recognize speech and translate it into text, SplitCam will act as a virtual webcam and for capturing a selected area from the desktop (or browser), VirtualAudioCable is necessary for microphone emulation, and text editor to write / save text.

For those who want to work with video not through a browser, they can download it to their computer (using special programs or websites) and work, for example, via KMPlayer or any other player.

We proceed to the implementation.

I will not describe the software installation, but I'll start right away with the settings.

First of all, we set up VirtualAudioCable, for this we launch VirtualAudioCable-> Controlpanel from the start. In DriverParameters, opposite Cables, specify 2 and press Set. Then select the first cable and tick the checkbox next to VolumeControl and click Set.

You should have this:



Close the Controlpanel by clicking on Exit.

Go to Start -> Control Panel -> Hardware and Sound -> Sound



Go to the Playback tab and select Line 1, press the “Default” button, until we close the sound settings, they will still be useful to us for testing.

Now let's set up a virtual cable, and for this we go to Start-> VirtualAudioCable-> Audio Repeater (MME) and configure it the same way as I have shown in the screenshot and click Start.



In the WaveOut field, you need to select your audio output device (speakers, headphones, etc.), I recommend to run some movie or music to check the output. If you hear a sound, then everything is set up correctly.

You can check who returned to the sound settings

If during playback you have about the same as I have in the screenshot, then you have correctly adjusted the sound output.



Turning to the Record tab, Line 1 should move as well and the virtual microphone setup is now complete.



Now we need to get a video stream, run SplitCam.

Click on the Sources -> Screen -> tab and select a screen area, as shown in the screenshot.



Direct, expand or narrow the appeared area on the announcer's face from a video clip (for example from YouTube).



After selecting the area, press the close button and you can minimize SplitCam.

Virtual camera and virtual microphone are configured, the area is selected, now you need to configure RealSpeaker and translate speech into text.

Launch RealSpeaker, click on the six-needle and indicate the settings as shown in my screenshot, namely Microphone: Line 1, Camera: SplitCam VideoFilter. Click OK.



If everything is set up correctly, you will see the following:



Now click on video playback, then on RealSpeaker to start the process of receiving video and audio stream, and move the cursor to a text editor where you need to write the received text.

I recommend still saving the video and launch it for example in KMPlayer and play literally 5-10 seconds, then pause the translation of the cursor back into a notebook for writing text. RealSpeaker sends and receives data through Google, respectively, there is a delay in recognition, and if the announcer quickly dictates the text, the recognition is sometimes curved. Just watch the keyboard layout, as the program stupidly presses the keys and does not translate the layout.

I think everything, try and go to our forum .