For those who don't know me my graduate studies (MSc 🇨🇴 & Phd 🇸🇦) centered around video understanding. After discussing with young lads, I figured it out that there might not be a encompassing defition of video. Thus,
WIP
Around the middle part of my PhD, I got to know about the Pythonic scientific kit (scikit) toolset for dealing with videos, scikit-videos.
I was still inmature, and disregarded it as I found it slow and just a nice wrapper around my go to FFMpeg. The latter is perhaps the most popular video processing tool of the world.
🙋🏽♂️ Why did I put my eyes on sci-kit video?
Perhaps, I was interested on using motion vectors to speed up conv-nets. Unfortunatly, I could not retrieve a cool tutorial about video compression that I read back in ~2017. But, I found the FFGligtch which looks cool to watch after harvesting season 🍄.
🙋🏽♂️ Haven't I taught anything yet?
My bad 😞, I'm sorry for wasting your time.
Lemme depart with some dots that I connected but I haven't confirmed.
PyAV is a Pythonic wrapper for the AV decoding library.
Think about AV as the audio-visual layer behind FFmpeg.
If you are a (genuine) computer vision scientist (de los de la mata 🇪🇸), libjpeg