Audio RADAR-pRMT
Audio plugin for the RADAR passive remote monitoring app. It uses openSMILE to process audio from the phone's microphone, extracts features from it, and sends those features.
Installation
To add the plugin code to your app, add the following snippet to your app's build.gradle file.
dependencies {
implementation "org.radarbase:radar-android-audio:$radarCommonsAndroidVersion"
}
Add org.radarbase.passive.audio.OpenSmileAudioProvider to the plugins variable of the RadarService instance in your app.
Configuration
To enable this plugin, add the provider opensmile_audio to plugins property of the configuration.
| Parameter | Type | Default | Description |
|---|---|---|---|
audio_duration | int (seconds) | 15 | Length in seconds of the audio recording when it started. |
audio_record_rate | int (seconds) | 3600 | Default interval between two consecutive audio recordings. |
audio_config_file | string (filepath) | "ComParE_2016.conf" | Path to openSMILE configuration file. |
Config files are addressed relative to the path src/main/assets/org/radarbase/passive/audio/opensmile.
Please note: The actual audio recording duration is stated here:
src/main/assets/org/radarbase/passive/audio/opensmile/shared/FrameModeFunctionals.conf.inc
However, the audio_duration parameter from the RadarConfiguration parameters only tells after
which time period a termination signal should be sent to the openSMILE feature extraction thread.
So audio_duration and frameSize and frameStep in FrameModeFunctionals.conf.inc must all be set
together correctly such that all values correspond to each other.
This plugin produces data for the following topics: (types starts with org.radarcns.passive.opensmile namespace)
| Topic | Type | Description |
|---|---|---|
android_processed_audio | OpenSmile2PhoneAudio | Raw base-64 encoded data, according to the provided config file. |
License
The radar-android-audio plugin includes a modified copy of OpenSMILE, and it is subject to the src/main/jni/opensmile-2.3.0/COPYING file. In addition, Please cite openSMILE in your publications by citing the following paper:
Florian Eyben, Martin Wöllmer, Björn Schuller: "openSMILE - The Munich Versatile and Fast Open-Source Audio Feature Extractor", Proc. ACM Multimedia (MM), ACM, Florence, Italy, ISBN 978-1-60558-933-6, pp. 1459-1462, 25.-29.10.2010.
Please note that this plugin may only be used for personal or academic purposes, or for commercial research not directly associated with product development and with the primary purpose of publishing and sharing results with the academic world. For the exact terms, please refer to the src/main/jni/opensmile-2.3.0/COPYING file.