How to use VOICEDAR LAB

General Guide To VOICEDAR LAB

1. Go to Upload Files to upload your data sets

2. Train your own model and perform visualisation on data: Training & Visualisation

1. Go to Compare Models to compare your models and see a detailed report on their performances.

How To Upload The Sound Files

1.  Use the Electron App to locally upload sound files of recorded speech. To download the Electron App, please go here.

2.  Navigate to Upload Files and upload the CSV files created in the previous step. Please make sure to add a comment and (if available) label for each upload.

Data Visualisation

1.  Navigate to Training & Visualisation and select two (!) files badges from your uploads.

2.  Use the 'Visualisation' button and choose which features extracted from the uploaded files you would like to analyse.

2.  Select the visualisation of your choice (BoxPlot, OLS, 3D) and see the detailed analytics report with statistical analysis of

Train Your Own Model

1.  Navigate to Training & Visualisation and select two (!) files badges from your uploads.

2.  Use the 'Train Your Model' button to train a new classification model which can later be retrieved for classification purposes.

Conduct Classification

1.  Navigate to Training & Visualisation and select one (!) files badge from your uploads..

2.  Use the 'Prediction' button to classify (i.e., predict the correct label) for a given file badge using a pre-trained model.

Compare Models

1.  Navigate to Compare Models and select one or more of the models created.

2.  Use the 'Compare models' button to analyse the performance of each model relative to each other and share with your colleagues (use PDF button at the bottom right of your screen).