Acciona Manifesto

  • Client:
    • Acciona
  • Studio:
  • Role:
    • Design & Development
  • Tools Used:
    • Stable Diffusion 1.5
    • ESRGAN
    • Parseq
    • Adobe AfterEffects
  • Creation Date:
    • June 2023

In the early days of the generative AI boom we utilized AI to successfully craft a video that showcases the company’s values, mission, and vision.

The entire process of image generation and processing for this project was carried out using open-source software like Stable Diffusion 1.5.

‍

Firstly, we carefully selected the most relevant images based on the script provided by Acciona. We then applied filtering and utilized OpenAI’s CLIP AI to analyze the content of the images, aligning them with references and prompts.

Next, we sequenced the values to achieve a coherent generation of images through Stable Diffusion, generating hundreds of sequences based on a reference image and a target prompt as input.

‍

To ensure the highest visual quality, we upscaled the images using ESRGAN models. As a final step before editing, we generated interframe smoothing information to add temporal coherence.

A total of 43,111 images have been generated, amounting to a total of 580 GB of data and over 166 sequences.

No items found.