This is a simple video-to-slide converter application that aims to obtain slide images (or pdf) given slide or lecture videos.
This is highly useful when one wishes to have a video lecture(with or without animations) in the form of slides – either a ppt or pdf. However, more often than not, slides are not provided when such video lectures are hosted on platforms like YouTube. This project aims to build a robust application that can convert video lectures into corresponding slides using techniques such as basic frame differencing and statistical background subtraction models such as KNN or GMG.
Notes:
- It will take some time to complete (about half of the original video length), so stay tuned!
- If the YouTube video link doesn't work, you can try again later or download the video and process it locally
- For slide videos with full background animation similar to this video, the algorithm cannot extract the right frames after tthe animation ends. In that case, using Frame Differencing (
--type Frame_Diff) will yield better results, but the results are still decent. - You may want to trim the video before processing it for better results.
-
Install OpenCV If you already have opencv installed, skip this step
pip install opencv-contrib-python
-
Install other packages
pip install -r requirements.txt
-
Install gradio (optional) If you don't use the GUI version, skip this step
pip install gradio>=4.0
usage: video_2_slides.py [-h] [-v VIDEO_FILE_PATH] [-o OUT_DIR] [--type {Frame_Diff,GMG,KNN}] [-hf {dhash,phash,ahash}] [-hs {8,12,16}]
[--threshold {90,91,92,93,94,95,96,97,98,99,100}] [-q QUEUE_LEN] [--no_post_process] [--convert_to_pdf]
This script is used to convert video frames into slide PDF.
optional arguments:
-h, --help show this help message and exit
-v VIDEO_PATH, --video_path VIDEO_FILE_PATH
Path to the video file, video url, or YouTube video link
-o OUT_DIR, --out_dir OUT_DIR
Path to the output directory
--type {Frame_Diff,GMG,KNN}
type of background subtraction to be used
-hf {dhash,phash,ahash}, --hash-func {dhash,phash,ahash}
Hash function to use for image hashing. Only effective if post-processing is enabled
-hs {8,12,16}, --hash-size {8,12,16}
Hash size to use for image hashing. Only effective if post-processing is enabled
--threshold {90,91,92,93,94,95,96,97,98,99,100}
Minimum similarity threshold (in percent) to consider 2 images to be similar. Only effective if post-processing is enabled
-q QUEUE_LEN, --queue-len QUEUE_LEN
Number of history images used to find out duplicate image. Only effective if post-processing is enabled
--no_post_process flag to apply post processing or not
--convert_to_pdf flag to convert the entire image set to pdf or notIf you want to manually remove some images before generating final PDF file, you can use the convert_to_pdf.py script later to convert the entire image set to pdf
usage: convert_to_pdf.py [-h] [-f FOLDER] [-o OUT_PATH]
This script is used to convert video frames into slide PDFs.
optional arguments:
-h, --help show this help message and exit
-f FOLDER, --folder FOLDER
Path to the image folder
-o OUT_PATH, --out_path OUT_PATH
Path to the output PDF file. If None, the image
directory will be used to store the output file.gradio app.pyThe application will be available at http://localhost:7680
| Video file | Output |
|---|---|
| https://www.youtube.com/watch?v=bfmFfD2RIcg | output_results/Neural Network In 5 Minutes.pdf |
| https://www.youtube.com/watch?v=rQijrDj1wCQ | output_results/What is Docker Container Docker explained in 5 minutes.pdf |
| https://www.youtube.com/watch?v=YxlDoz_P4kc | output_results/Introduction to Stable Diffusion.pdf |
| sample_vids/react-in-5-minutes.mp4 | output_results/react-in-5-minutes.pdf |
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent