A web page as a demonstration/ testing site for my understanding of CNN or Convolutional Neural Network in Machine Learning.
The web app follows the notebook by Shivam Burnwal on Kaggle: https://www.kaggle.com/code/shivamburnwal/speech-emotion-recognition
- Accepts audio recording from the web.
- Feeds the audio into a pre-trained model that predicts one of the following emotions (multiclassification):
- neutral
- calm
- happy
- sad
- angry
- fearful
- disgust
- surprise
- Asks for the intended emotion for feedback.
- Forms a sample that includes the intended emotion as label for that sample.
-
Install the requirements
$ pip install -r requirements.txt -
Run the app
$ streamlit run streamlit_app.py