Audx - Android Real-time Audio Denoising Library

A streaming, real-time audio denoising library for Android with Voice Activity Detection (VAD). It wraps a native C++ library (audx-realtime) with a clean, easy-to-use Kotlin API.

Features

• Real-time Noise Suppression: Based on the excellent Xiph.Org RNNoise algorithm.
• Streaming API: Process audio in chunks of any size. The library handles all internal buffering.
• Support for Any Sample Rate: Automatically resamples from any standard input rate (e.g., 8kHz, 16kHz, 44.1kHz) to the required 48kHz and back.
• Voice Activity Detection (VAD): Each processed chunk includes a probability score for speech presence.
• Performance Statistics: Collect detailed metrics on CPU usage and speech activity.
• Fluent Builder API: A simple, chainable builder for easy configuration.
• Lightweight & Performant: Efficient C++ core with no external dependencies outside of the Android NDK.
• Zero-Copy API: A new processAudio(ByteBuffer) API for high-throughput scenarios, offering true zero-copy processing when used with a DirectByteBuffer.

Performance

Version 1.1.2 introduces significant performance improvements, including pre-allocated memory buffers and cached JNI lookups, resulting in:

• ~25-35% faster processing per frame.
• ~20-25% lower CPU usage.
• Zero memory allocations on the audio processing hot path.

For high-performance use cases, the new zero-copy API is recommended. See the Quick Start Guide for an integration example.

Getting Started

For a detailed guide on integrating Audx with AudioRecord in a modern Android app, please see the Quick Start Guide.

Here is a minimal example:

// 1. Build the denoiser instance
val denoiser = AudxDenoiser.Builder()
    .inputSampleRate(16000) // Match your source's sample rate
    .onProcessedAudio { result ->
        // This callback runs on a background thread.
        // 'result.audio' contains the denoised audio.
        Log.d("Audx", "VAD Probability: ${result.vadProbability}")
    }
    .build()

// 2. Process audio chunks as they arrive (e.g., from AudioRecord)
denoiser.processAudio(myAudioChunk)

// 3. At the end of the stream, close the denoiser to release resources.
// This is a blocking call that performs a final flush before cleaning up.
denoiser.close()

Installation

1. Add the JitPack repository to your project's settings.gradle.kts file:

dependencyResolutionManagement {
    repositories {
        google()
        mavenCentral()
        maven { url = uri("https://jitpack.io") }
    }
}

2. Add the dependency to your module's build.gradle.kts file:

dependencies {
    implementation("com.github.rizukirr:audx-android:1.1.2")
}

Always check the Releases page for the latest version.

Documentation

• Quick Start Guide: A detailed, step-by-step integration guide.
• API Reference: A complete reference for all classes, methods, and constants.

Requirements

• Min SDK: 24 (Android 7.0)
• Audio Format: 16-bit signed PCM, Mono.

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Support This Project

If you find this project helpful, consider ☕ Buy Me a Coffee