HOW YOLO V3 WORKS?

preview_player
Показать описание
In this video I will focus on how Yolo algorithms(mainly yolov3) work.
So what is happening between feeding the image to the network and getting the detections.
I will also share with you the tool that I created that is an interactive notebook,
that allows you to explore this alogrithm on your own.
Link to the repository:

Table of context:
0:00 Intro and installation
3:11 How back-bone feature extractor works
11:30 From feature maps to detection
23:47 How does YOLO picks the right detections
31:13 Tips for better detection
37:49 Which implementation of YOLO to use
Рекомендации по теме
Комментарии
Автор

Excellent explanation. I really like your style and the level you picked.

xlxlxl
Автор

Amazing, mostly all the videos are just yolo in 5 minutes. But this was really helpful
for me, thanks a lot.

skaayth
Автор

I'm a beginner in object detection, this is really helped me lot....expecting more for people detection and gender identification

harikrishnasamy
Автор

amazing friend have never seen an tutorial like this i am very greedy now to copy this tutorial to my channel you have given me the wisdom of yolo so i am just admiring it . copying is an shame full act so i will surely subscribe and tell my subscriber to follow this cool guy on YouTube you are awesome i am so excited thank you for this wonderful tutorial

ranam
Автор

Thanks a lot! One of the best explanation ❤

nemonovich
Автор

Oh this Video helped me a lot to understand, Thanks!!!

Dayisa
Автор

So much good stuff in this video, it's just amazing.
Thank you a lot! :D

cristiancorches
Автор

awesome explanation. really appreciated.

ytohustler
Автор

thanks you so much! This was really helpful

hoangtung
Автор

at 12:06 you mentioned 255( 3 predictions 4+1+80). Could you explain a bit how did you get these 3 predictions after performing convolution over your feature maps..

vipingautam
Автор

Thank you for a wonderful video! I have an equation here in the last convolutional for the detection layer you have 1*1*18 how does it come? thank you

hazemabolholl
Автор

Hey
Nice video. Really informative.
You have not added the paint image to the GitHub repository.
Please add

cvbasics
Автор

Great. Just there is no comparison of V3 vs. V4... still, very good!

arielherold
Автор

Hello man i hope you are okay im asking
Is there any way to increase the accuracy of yolo v4 maybe we add more layers or something else?

zubairsk