Intro

To make a video and a Handy move, we need to have a Script. The process of making a Script is not straightforward. We, in Ohdoki , has developed a scripting pipeline that utilizes computer vision and AI to simplify the job. However, videos have a lot of different scenes and angles, and it's not always perfect. Our automation level is now at around 60-95%, depending on the video using Machine Learning (ML) and Computer Vision (CV). Videos captured with our guidelines will be on the upper range of the scale.

Scripting process

We predefined the scripting pipeline with the following steps:

  1. You upload videos or videos to our server using FTP (simple upload through FilleZilla or similar)
  2. You check our admin tool to verify the uploads and set some additional parameters
    1. Unique id. This can be set automatically
    2. Set priority. It might be that you want some videos to skip the scripting queue
    3. Verify the upload
  3. The video is added to our scripting task list
    1. We might have priority clients that pay us for the scripting job. Their videos will always go in front of the task queue.
  4. Our ML+CV algorithms run through the video and make an initial script.
  5. The scripting task is assigned to one of our scripters.
  6. The scripter goes over the video and the automated script. Where the automation has not been able to script, they will manually go in and script themself.
  7. The script goes to QA and is published.
  8. If you opt-in, then the script is published on our site ivdb.io.

Pipeline admin

How to start using our scripting pipeline

Reach out to us first and tell us about your site. Then follow this guide on how to upload videos:

Step-By-Step Guide to Upload Videos to Our Scripting Pipeline

Scripting tool

We provide a scripting tool for our community. Just so you know, this tool has a license, so you can only use it to create free and publicly available scripts. Are you unsure if that applies to you? Reach out and ask us.