FineTuned

Hack the North 2020++ project! Devpost Submission | YouTube Demo

The memory-making machine that tunes your photos to the song of the moment.

Demo

Getting Started

Prerequisites

This project was built using Python3, which can be installed here.

Installation

1. Clone this repository onto your local device
2. The files spotify.py and genius.py both require API keys.
3. A Spotify API key can be acquired at this link. Copy both the client_id and secret into cid = '' secret='', respectively, in spotify.py
4. An API token for Genius can be acquired pursuant to these instructions. This should be copied into API_TOKEN = '' in genius.py
5. A token is also needed for Microsoft Azure. More information can be found here. The subscription key and endpoint should replace subscription_key = "" and endpoint = "" in genius.py

Deployment

1. Run python server.py from command line in the main folder.
2. The web app should open locally at http://127.0.0.1:5000/

Usage

A user may choose to explore the home page briefly. To upload photos, they can either navigate from the home page or using the Navbar to get to http://127.0.0.1:5000/start. Following this, the user follows through a fairly self-explanatory procedure of selecting a song before viewing how the photos are edited accordingly.

Known Bugs

Due to some caching issues, when the same image is loaded in multiple times to be viewed properly on screen in their edited form.

Future Steps

Some immediate next steps we could take involve making a more user-friendly interface for users to download their edited photos, as well as further working through the details of the tuning algorithm.

Long term, AI could have play a potential role in the photo-editing process. Currently, the coefficients used in photo-editing are hand-picked; however, with more sample data, it is possible that neural networks could make the process more consistent.

Built With

This project was built with:

• Flask - Web Framework
• Bootstrap Navbar - Bootstrap components
• Jinja - Web Template Engine

Acknowledgements

Inspiration was drawn from:

• Osvaldas Valutis and his form input styling

License

This project is licensed under the terms of the MIT license and is available for free use.