Skip to content

Latest commit

 

History

History
35 lines (23 loc) · 1.18 KB

README.md

File metadata and controls

35 lines (23 loc) · 1.18 KB

Flask Audio Processing Application

This is a Flask application that allows users to easily demo senseflosser without using CLI after downloading the repo and installing dependencies.

Features

  • Upload audio files
  • Process audio files using different models
  • Play processed audiofile

Installation

  1. Follow directions in the primary repo README

Usage

You can start the flask app in a couple of ways. Assuming you are in the base directory of the repo on your computer:

cd flask
flask run

Then, open a web browser and navigate to http://localhost:5000 or http://127.0.0.1:5000.

Notes:

  • Model selection is based on what model files (.h5) are in senseflosser/models.
  • Max duration is determined once a file is chosen and is not limited by the timeframe used to train the model (i.e. 15s_audio_autoencoder.h5 can process more than 15 seconds of audio)

Endpoints

  • GET /: Render the main page of the application.
  • POST /: Process the uploaded audio file and return the result.
  • GET /output/<filename>: Serve the processed audio file.
  • POST /upload: Upload an audio file and return the maximum duration.