kaiidams / NeMoOnnxSharp

Text-to-speech and speech recognition, VAD with NVIDIA NeMo and ONNX Runtime for .NET Core.
Apache License 2.0
18 stars 2 forks source link

Godot recording settings #8

Closed GeorgeS2019 closed 2 years ago

GeorgeS2019 commented 2 years ago

Under Godot Project General settings, the sampling rate is set to 16000;

using an available tutorial => https://godotengine.org/asset-library/asset/527 Using the saved wav file as input, I am unable to get decent Speech Recognition results.

Is there something you would advice to improve the consistency of speech recognition using NeMoOnnxSharp? image