Skip to content

Latest commit

 

History

History
82 lines (57 loc) · 3.73 KB

README.md

File metadata and controls

82 lines (57 loc) · 3.73 KB

Kinetic Typography Diffusion Model

Instruction for the dataset construction


We provide detailed instructions on how to construct the kinetic typography dataset. Due to IP licensing issues with the templates, we also provide source code for post-processing and video rendering. After downloading the original template files from the official webpage and following these instructions, you can easily create the kinetic typography dataset. Details are as follows.

Preliminaries


Examples of our KineTy dataset. You can download them with this link.

To construct the dataset, please ensure you have Adobe After Effects installed, as our dataset is created using Adobe After Effects templates. Additionally, you will need other programs related to Adobe After Effects.

Adobe After Effect

You can download the software from here: Adobe After Effect.

Nexrender

We use only 'nexrender-cli-win64.exe' from Nexrender Installation link. This is for Windows and we use version 1.46.6. This program enables the use of Adobe After Effects through programming. For more information about this, you can find it here: Nexrender

Templates

We will provide you with templates, but you can download more Adobe After Effects templates from the site below.

Settings for Running Code

📝 Text

Write down the words you want to print out in the video to word.csv

🎥 Video setting

Edit rendering information such as resolution and frame number in basic_rendering_info.json. You can add or change the other elements in reference to Nexrender.

📁 Template setting

Record the below information in template_info.csv.

  • id: just a number of the template
  • template_name: template file name
  • composition_name: a composition name what you want to render
  • composition_text: the composition name where the text layer is
  • layer_text: a text layer name which has a 'source text'

Video Rendering

For using nexrender:

pip install ffmpeg

Then, set your path to dataset construction folder in render_set.py:

# Path to your project folder
basedir = "~~~/dataset_construction"

Lastly, designate your path to nexrender software and aerender software in render.py:

# Path to your nexrender-cli-win64.exe
nexrender_cli_path = "~~~\\nexrender-cli-win64.exe"

# Path to your Adobe After Effects aerender.exe
aerender_path = "~~~\\aerender.exe"

Prepare rendering information with each video and render the template:

python render_set.py    # Save the rendering information 
python render.py        # Render the template into a video