Things to Note When Using Google Recognition
- You need to fill in a network proxy, otherwise you cannot connect.
- Google's speech recognition is not very robust and cannot distinguish and return punctuation.
- Suitable for audio recognition with clean background sound and clear and accurate human voices.
How to Use
Select GoogleSpeech from the mode drop-down box in the software interface. When this item is selected, there is no need to select the model and segmentation method.
Advantages and Disadvantages
Advantages: No need to download models, saves system resources, simple to use.
Disadvantages: Requires a proxy, slightly worse results.
Proxy Issues
When using services such as Google Gemini, you must use a proxy for well-known reasons. The general form is http://127.0.0.1:number port number. If you have confirmed that you are using a system proxy and can access it in your browser, but you 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 pop-up Run window, and then click OK.
If the pop-up settings panel is similar to the figure below, it means that the system proxy has been set correctly, and there is no need to fill in the "Network proxy address" in the software.
