Skip to content

samanthajmichael/ml_project

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Match Cut AI: Revolutionizing Movie Trailer Creation

Project Overview

Match Cut AI is an innovative project that aims to revolutionize the creation of movie trailers using machine learning techniques.

Our goal is to automate and enhance the process of match cutting, a crucial technique in creating compelling trailers.

The Challenge

Match cutting, is a time-consuming process that can take days, especially when working with 2000+ frame shots in a single movie.

Our project addresses the challenge of efficient match cutting in video editing, specifically for movie trailers.

Our Vision

We envision a future where multiple trailers of the same movie, each tailored to different audiences, can be created in a fraction of the time it takes today.

Technical Approach

Our approach leverages cutting-edge machine learning techniques:

  • Object Detection: Using tools like TensorNet and torchvision to train models on movie frames.
  • Contrastive Learning: Teaching the model to differentiate between similar and dissimilar inputs, capturing subtle visual similarities that make for compelling transitions.

Key Features

  • Automated identification of potential match cuts in thousands of frames
  • Ability to create multiple, audience-tailored trailers efficiently

Initial Dataset

For our initial training, we plan to use the black and white movie "To Kill a Mockingbird" to train a base model. As we progress, we'll enhance and tune the model, potentially expanding to more diverse datasets.

Project Goals

  • Develop a machine learning model capable of identifying potential match cuts
  • Significantly reduce the time required for match cutting in trailer creation
  • Enable the creation of multiple, diverse trailers for the same movie

Future Directions

  • Expand the training dataset to include a wider variety of movies
  • Incorporate user feedback to continually improve the model's performance
  • Explore additional applications of the technology beyond movie trailers

About

Machine Learning project on match cutting in video editing

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •