Match the given frames from a lecture video to the original slides.
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.
All the dependencies are stored in
requirements.txt
file.
Following command can help install those on Linux:
pip install -r requirements.txt
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 inresults.txt
according to the present naming convention. This script can be changed according to the requirement.