VideoThat

Automagically Generate a Music Video for Your Input Audio

2nd place DataHack winner 🥈, September 2019

VideoThat will automagically generate the perfect video compilation that fits impeccably to the theme song of your choice. The algorithm’s default choice is to create a video clip from Taylor Swift’s videos to an Arctic Monkey song.

This project was created as part of DataHack 2019 Israel, read all about it in our blog post 👾

---

Installation

Fork VideoThat repository to your file system

System requirements

Ubuntu/debian

sudo apt install ffmpeg

macOS

brew install ffmpeg

Install requirements file

pip3 install -r requirements.txt

Install the newest version of youtube-dl

pip3 install --upgrade youtube-dl

---

Code pipeline

Create Data-Base (db) of video clips

To start downoading all of the videoclips and create a data-base, open the terminal go to the repository location and write

note: the following command will download ~1.5GB to your directory and can take a while

./download-db.sh

Detect all of the scene cuts from the db you just created

./scenedetect_all.sh

Run create_scene_db.py to create a .pkl df unified file of all the scenes from the db (this creates scene_db.pkl and song_db.pkl)

python3 create_scene_db.py

Run choose_scenes.py to run the algorithm which chooses taylor swift scenes from the db to match to a song (creates chosen_scenes.pkl)

python3 choose_scenes.py

Run build_video.py to create the video file from the chosen scenes

python3 build_video.py

The video videoThat.mp4 has been successfully created and added to your directory

---

Code structure

---

That's Us

Yael Daihes GitHub LinkedIn

Yaara Arkin GitHub LinkedIn

Orian Sharoni GitHub LinkedIn

Roee Shenberg GitHub LinkedIn

Dalya Gartzman GitHub LinkedIn

---

P.S.

Our prise money was donated to TAMI - Tel Aviv Makers Isocratic hackerspace