/ Blog

Taxo Cat Teaser - First AI Gen Assisted video

selectable region in runway app

I've always dreamed pretty vividly awake or asleep. I've often created mind movies that play when I hear familiar songs. I'm excited that I think I finally found a way to start sharing those moments with others. ✨

In the past I've worked a lot with the Adobe suite creating graphics and videos. Adding AI assisted tools looks like its going to supercharge my efforts. I've created a short little video has a proof of workflow and capabilities.

This is nothing new as a ton of people have already found this flow. Iit's new though for so I'm sharing here anyways.

Here's the flow:

  1. Decide to make art. However, you want to make it. Just express yourself. Even it may seem odd, funny or like you're the only one who will "get it".
  2. Pick a vision to share
  3. Create the source image
  4. Use the image with accompanying text to create a video segment
  5. Continue creating keyframes until you have something, it won't be perfect or likely even match the original vision.

Creating keyframes

I can pencil or marker sketch preety decently when trying to illustrate a vision but I don't often spend time or have great skills finishing art enough to my liking. I used to remix and enhance images a ton in Photoshop.

The supercharge hack here is to try Scenario. I'm really impressed with it's gallery of AI models and ability to maintain character appearance from image to image.

a screnshot of the Scenario app

It can be challenging to land on a winner image with your prompt and every image has required work within Photoshop to clean up extra body parts or odd artifacts. For quickly "sketching" out ideas though it's amazing and a real boost to productivity.

I'll make tutorials on it someday, if I ever feel like I know what I'm doing beyond typing and rolling the dice to get lucky. 🎲

Creating video segments

I jumped in with Runway and have been enjoying the journey so far. Features like, selectable regions, keyframe-based generation (first or last), and lip-synching drew me along with the amazing output that it's generating.

This is the most nascent practice for me and generating videos with images has been amazing, ridiculous, and really exciting. I mean look at some of these early attempts below.😂

mishaps attempting render the cat video

The teaser video

It's a teaser video, now that I've seen it. I thought this was just an experiment but there is a whole story now. I'm looking forward to bring that and the accompanying music to life that goes with the story. ♭🎞️