Skip to content

A tool to match given lecture video frames to the original slides.

Notifications You must be signed in to change notification settings

sajalasati/Slides-Recognition-Tool

Repository files navigation

Slides Recognition Tool

Match the given frames from a lecture video to the original slides.

Language Module

Made as a part of Digital Signal Analysis Course (DSAA) Final Project alongwith two other team members Arnav Juneja and Preet Thakkar. Given two sets of images stored in folders named - Slides and Frames - the script will output the corresponding matching ppt slide for each frame.

Dependencies

All the dependencies are stored in requirements.txt file.

Following command can help install those on Linux:

pip install -r requirements.txt

Usage

The script.py contains the code to classify the frames images.

  • Run the following command on terminal:
python3 script.py <path-to-frames-folder> <path-to-slides-folder>
  • This generates a result.txt file, containing a mapping of each frame to its corresponding slide.
  • While the code runs, it also displays running accuracy on the data processed so far (as we have kept the names of images in some particular manner for now).
  • check.py script can check the accuracy of the output stored in results.txt according to the present naming convention. This script can be changed according to the requirement.

About

A tool to match given lecture video frames to the original slides.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published