Denoise

The RNNoise plugin enhances audio quality by removing background noise from your audio input, resulting in improved speech-to-text (STT) accuracy and better overall audio processing performance.

RNNoise is a real-time noise suppression library powered by a recurrent neural network that intelligently filters out environmental noise such as air conditioning, computer fans, and other stationary background sounds while preserving the clarity and quality of speech.

Installation

Install the RNNoise plugin for denoising in VideoSDK Agents package:

pip install "videosdk-plugins-rnnoise"

Importing

from videosdk.plugins.rnnoise import RNNoise

Example Usage

from videosdk.plugins.rnnoise import RNNoise
from videosdk.agents import CascadingPipeline

# Initialize the RNNoise Plugin
rnnoise = RNNoise()

# Add Denoise Plugin to cascading pipeline
pipeline = CascadingPipeline(denoise=rnnoise)

It also works with RealTimePipeline.

Example Usage in RealTime Pipeline

from videosdk.plugins.rnnoise import RNNoise
from videosdk.agents import RealTimePipeline

# Initialize the RNNoise Plugin
rnnoise = RNNoise()

# Add Denoise Plugin to realtime pipeline
pipeline = RealTimePipeline(denoise=rnnoise)

Benefits

• Enhanced STT Accuracy: Cleaner audio input leads to more accurate speech-to-text transcription
• Real-time Processing: Processes audio streams with minimal latency for seamless user experience
• Intelligent Noise Reduction: Effectively removes background noise while preserving speech clarity

Additional Resources

The following resources provide more information about using RNNoise with VideoSDK Agents SDK.

• Python package: The videosdk-plugins-rnnoise package on PyPI.

• GitHub repo: View the source or contribute to the VideoSDK RNNoise plugin.

• RNNoise project: The open source RNNoise library that powers the VideoSDK RNNoise plugin.