Project: StoryBoard-Generation

Introduction:

This project aims to automate the process of transforming textual descriptions of advertisement concepts and assets into detailed storyboards. The project leverages machine learning, natural language processing, and computer vision technologies to generate visually compelling storyboards that encapsulate the essence of proposed ad campaigns. This README provides an overview of the project, its objectives, data sources, and tasks involved.

Business Objective:

The objective of this project is to streamline the advertising production process by automating the conversion of textual advertisement concepts and assets descriptions into visually engaging storyboards. By harnessing advanced machine learning algorithms, the goal is to significantly reduce the turnaround time for creating ad campaigns while enhancing creativity and efficiency.

Background & Context:

The project is inspired by recent advancements in machine learning and digital advertising technologies. It aims to bridge the gap between textual concepts and visual storytelling, enabling advertisers to create detailed and dynamic content that resonates with viewers. The provided resources offer insights into the technical terms, creative aspects, and industry context relevant to the challenge.

Data:

The project utilizes various datasets, including assets folders containing images, sample concepts with assets and size descriptions in JSON format, and storyboard examples. These datasets serve as inputs for generating visual and textual assets, composing advertisement frames, and building the final storyboard.

Tasks:

The project is divided into three consecutive major tasks, each building upon the insights and outputs of the previous one:

1. Task 1: Image and Text Generation

   • Generate images and text based on provided descriptions.
   • Explore methods for producing high-quality visual and textual content aligned with advertisement concepts.
   • Implement image generation techniques considering aspect ratio, quality, and specificity.

2. Task 2: Image Composition

   • Organize and compose generated images into advertisement frames.
   • Determine optimal placement, size, and orientation of assets within frames.
   • Apply creative design principles to assemble coherent ad frames.

3. Task 3: Building the Storyboard

   • Represent and visualize the storyboard using generated frames.
   • Arrange frames in a sequence that conveys the progression of the ad's narrative.
   • Incorporate variations for ads with branching narratives or multiple user interaction paths.

Contributors:

• Mubarek Hussen
• Basilel Birru

License:

This project is licensed under the MIT License.