TomatoFT / Speech-Emotion-Recognition

Recognizing the emotion of an speech audio with ConvNet, SVM, KNN and Logistic Regression

Speech Emotion Recognoiton

This is the repository for Final Project of DS102 (Statistical Machine Learning) of University of Information Technology (UIT)

First you need to install Python first

The whole training process is on Google Colab Notebook, just Ctrl F9 to make all things done

Then the colab will download your training models to your local machines, you will need Anaconda to set up Tensorflow more easily. Do the following steps to run the demo of these algorithms

Set up the environment to demo

The RecordAudio.py will record your voice and save it to output.wav. PredictEmotion.py will predict your Emotion through your audio.

Feel free to clone my code and i will appreciate if you update it to have better performances

Demo in Command Prompt

First clone the repository

git clone https://github.com/TomatoFT/Speech-Emotion-Recognoiton

cd Speech-Emotion-Recognoiton

Create Anaconda Virtual Environment And Install Dependencies

conda create --name TrialEnv

conda activate TrialEnv

conda install pip

pip install -r requirements.txt

Set up the demo of your Emotion Prediction of your voice

python RecordAudio.py

python PredictEmotion.py

Exit the Anaconda Virtual Environment

conda deactivate

Demo in Flask Web-app

Clone the repository and create Anaconda Virtual Environment and install dependencies same as Command Prompt Demo

python app.py

NOTICE: You must speak after clicking Recording