ERL: Emotion Recognition Library

ERL is a Python library that does emotion recognition through audio. With this library you can recognize emotions using audio signals, the text extracted from the audio or using both.

Installation

• Clone the source of this library: https://github.com/estefaaa02/ERL.git
• Install library: pip install -e .

Note:

• If using Windows download the PyAudio wheel for Windows and install it using pip install.
• If using Linux install the PyAudio library using pip install pyaudio
• If using Linux run apt install libasound2-dev portaudio19-dev libportaudio2 libportaudiocpp0 ffmpeg
• By doing pip install -e . you are installing the necessary dependencies for the library to work

Emotion classification example

from ERL import PredictionsModule

print("Predicción Audio: ", PredictionsModule.predict_emotion_audio_svm('ERL/data/es/f_ans002aes.wav'))
print("Predicción Texto: ", PredictionsModule.predict_emotion_text_cnn('ERL/data/es/f_ans002aes.wav'))
print("Predicción Bimodal: ", PredictionsModule.predict_emotion_bimodal('ERL/data/es/f_ans002aes.wav'))