ailia.audio samples

Use ailia.audio to perform MelSpectrum conversion of the input audio. Then classify the sound categories with the crnn model.

Requirements

• ailia SDK 1.2.11 and later
• Unity 2019.4.36f1 and later

Unity API

Document

ailia.audio C API

Setup

Clone this repository and open the project in unity.

git clone git@github.com:axinc-ai/ailia-audio-samples.git

This repository does not include ailia libraries. So you must get SDK and license.

Download ailia SDK trial version

Import ailia and ailia.audio libraries to plugin folder using copy_libraries.sh.

chmod +x copy_libraries.sh
./copy_libraries.sh

Place license file here.

- For Windows
 - Place ailia.lic in Assets/Plugins/x86_64/
- For Mac
 - Place ailia.lic in ~/Library/SHALO/

Run

You can run it by opening the sample scene (unity/Assets/AiliaAudioSampleScene.unity) and pressing Run. The inference result is displayed on the console.

C++ API

Document

ailia.audio C API

Setup

Copy ailia and ailia.audio libraries to cpp folder.

Make

Make with the following command.

cd cpp
export OSTYPE=Mac
make

Run

Execute with the following command.

./crnn_audio_classification

The inference result is displayed on the console.

input shape 176 128 1 1 dims 4
output shape 10 1 1 1 dims 2
predict result
dog_bark 0.868386

Python API

Document

ailia.audio Python API

Setup

Please clone ailia-models repository.

crnn_audio_classification

Run

Execute with the following command.

python3 crnn_audio_classification.py --ailia_audio

The inference result is displayed on the console.

 INFO crnn_audio_classification.py (91) : input: 24965__www-bonson-ca__bigdogbarking-02.wav
 INFO crnn_audio_classification.py (98) : Start inference...
 INFO crnn_audio_classification.py (109) : dog_bark
 INFO crnn_audio_classification.py (110) : 0.8681702017784119
 INFO crnn_audio_classification.py (112) : Script finished successfully.