+ 5
Q. What prerequisite knowledge will be required to create something like this?
How to create or try to create something like this? https://www.dropbox.com/s/c8kan4p8b9pw461/vid-20211206-wa0014.mp4?dl=0
5 Answers
+ 2
Kode Krasher Exactly!! This definitely isn't a one man project...
Shobhit :) I'm really ignorant when it comes to AI...this kind of code could perhaps even run on the browser if we could somehow stream video not just pictures idk but I think it's possible, see a few similar codes on the browser...one for facial recognition too
Then use a canvas to analyze the pixels streamed in from the video...pass those to a pre-trained and probably GPU accelerated Neural Network and use that output from the nn to mark/tag certain objects at certain parts of the screen....would probably have to run the nn multiple times to search for specific objects
All that sounds straightforward and all...but i tell you, no part will be as hard as training that nn...where do you even get data for stuff like that
It's definitely possible, not just in browser of course, but it's definitely no easy feat...
+ 6
Thank you so much Alfred Valentine and Kode Krasher for your valuable inputs. Thanks a lot.
+ 4