The core principle of video translation software is: to recognize text from the speech in the video, then translate the text into the target language, then dub the translated text, and finally embed the dubbing and text into the video.
The first step is to recognize text from the speech in the video. The accuracy of recognition directly affects the subsequent translation and dubbing.
Notes on Using Google Recognition
- A network proxy needs to be filled in, otherwise it cannot connect.
- Google's speech recognition function is not strong and cannot distinguish and return punctuation marks.
- Suitable for audio recognition with clean background sound and clear and accurate human voice.
How to Use
In the software interface, select GoogleSpeech from the mode drop-down box. When selecting this item, you do not need to select the model and segmentation method.
Advantages and Disadvantages
Advantages: No need to download models, saves system resources, easy to use
Disadvantages: Requires a proxy, slightly poor effect
Proxy Issues
When using Google Gemini and other services, due to well-known reasons, a proxy must be used. The general form is http://127.0.0.1:port number. If you have confirmed that you are using a system proxy and can access it in your browser, but don't know how to fill in the network proxy address, then execute the following command to confirm whether the system proxy is correctly enabled.
Press Windows key + R key, enter ms-settings:network-proxy in the Run window that pops up, and then click OK
If the settings panel that pops up is similar to the following image, it means that the system proxy has been set correctly, and you do not need to fill in the "network proxy address" in the software.
