Skip to content

Latest commit

 

History

History
58 lines (41 loc) · 3.53 KB

README.md

File metadata and controls

58 lines (41 loc) · 3.53 KB

Automated Storyboard Synthesis for Digital Advertising

Overview

This repository contains the implementation of a cutting-edge machine learning solution developed as part of the 10 Academy Cohort A Weekly Challenge: Week 10. The challenge focuses on automating the process of transforming textual descriptions of advertisement concepts and assets into visually compelling storyboards, thereby revolutionizing the digital advertising industry.

Business Objective

In response to recent advancements in machine learning, natural language processing, and computer vision, Adludio, a leading provider of online mobile advertising, aims to automate the end-to-end process of advertising production. By leveraging advanced machine learning algorithms, the objective is to significantly expedite the ideation and execution phases, enabling clients to swiftly launch their campaigns with minimal time and resource expenditure.

Folder Structure

  • .github/workflows: Contains GitHub Actions workflows for automating tasks such as testing and deployment.
  • data: Includes datasets and data preprocessing scripts.
  • generated_assets: Stores generated assets, including storyboards and individual frames.
  • logs: Contains logs generated during model training and evaluation.
  • models: Houses pre-trained models and model training scripts.
  • notebooks: Jupyter notebooks for exploratory data analysis and model experimentation.
  • screenshots: Screenshots of results and visualizations.
  • scripts: Utility scripts for various tasks.
  • tests: Unit tests for ensuring code functionality.
  • .env: Environment variables configuration file.
  • .gitignore: Specifies intentionally untracked files to be ignored by Git.
  • LICENSE: MIT license file.
  • Makefile: Makefile for managing project tasks.
  • README.md: Main repository documentation.
  • requirements.txt: Lists project dependencies for reproducibility.

Getting Started

To get started with the project, follow these steps:

  1. Clone the repository: git clone https://github.com/birehan/automated-storyboard-synthesis.git
  2. Navigate to the project directory: cd automated-storyboard-synthesis
  3. Install dependencies: pip install -r requirements.txt
  4. Explore the folders and scripts to understand the project structure and functionality.

Usage

  • Data Preparation: Use the scripts in the data folder to preprocess and prepare the datasets for model training and evaluation.
  • Model Training: Train machine learning and deep learning models using the scripts in the models folder. Experiment with different architectures and hyperparameters to achieve optimal performance.
  • Evaluation: Evaluate model performance using the provided evaluation scripts. Analyze metrics and fine-tune models as necessary.
  • Storyboard Generation: Utilize the provided scripts and notebooks in the scripts and notebooks folders to generate storyboards from textual descriptions of advertisement concepts and assets.
  • Testing: Run unit tests in the tests folder to ensure code functionality and reliability.
  • Logging: Monitor and analyze model training and evaluation logs stored in the logs folder.

Contributors

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements

Special thanks to the 10 Academy Cohort A team for organizing the weekly challenge and providing guidance throughout the project.