The Evolution of Convolution Neural Networks
Вставка
- Опубліковано 31 лип 2024
- From the one that started it all "LeNet" (1998) to the deeper networks we see today like Xception (2017), here are some important CNN architectures you should know. If you like the video, show your support with a like, and SUBSCRIBE for more awesome content on Machine Learning, deep Learning, Data Science and AI
MY EQUIPMENT (on a $350 budget)
Camera (GoPro Hero 5 Black + 32 GB Memory + Kit): goo.gl/V4542j
Microphone: goo.gl/BxBRcW
Pop filter: goo.gl/oQTQ8W
FOLLOW ME
www.quora.com/profile/Ajay-Ha...
REFERENCES
[1] LeNet-5 (the start of it all): yann.lecun.com/exdb/publis/pdf...
[2] Nice Blog post: towardsdatascience.com/neural...
[3] CNN Architectures: slazebni.cs.illinois.edu/sprin...
[4] ImageNet - The data that transformed AI research: qz.com/1034972/the-data-that-...
[5]Imagenet (main paper): www.researchgate.net/publicat...
[6] AlexNet: papers.nips.cc/paper/4824-ima...
[7] Difference between saturating & non-saturating nonlinearities: stats.stackexchange.com/quest...
[8] Top-1 accuracy Vs Top-5 Accuracy. What do they mean? stats.stackexchange.com/quest...
[9] OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks: arxiv.org/abs/1312.6229
[10] Network in Network (NiN) Architecture: arxiv.org/abs/1312.4400
[11] GoogleNet: arxiv.org/pdf/1409.4842.pdf
[12] R-CNN: arxiv.org/abs/1311.2524
[13] ResNet: arxiv.org/abs/1512.03385
[14] An Overview of ResNet and its Variants: towardsdatascience.com/an-ove...
[15] Xception: Deep Learning with Depthwise Separable Convolutions: arxiv.org/abs/1610.02357
Wow, just found your channel. It looks like a lot of really high quality content without competition on youtube! Great job - can't wait to watch them all. Keep it up and I'm sure you'll find continuing success boiiiiii
Thanks Sam! Hope you like the rest of the channel. This is the first video with the face cam. I'm going to use it in future videos too
The LeNet Family! I love this channel for always getting to fundamentals right away. You rock!
Thanks so much for the kind words and consistent support :)
Great to see you in the video, observed that the engagement is better with personal explanations. Thanks for sharing the knowledge.
"engagement is better with personal explanations" is exactly what I was going for. Glad to know someone else thinks so too. Thanks for the feedback, and glad you liked the video!
Yo man, this video content is so far the best, just loved the way you explained the intuition of why we go for deep models.
Very interesting, thanks!
You are very good. You teach me how to make such high quality videos and I will teach you how to smile :)
Keep up the great work.
Great review, must have taken a while, thanks for sharing that. Not sure if this is an intended effect, but you should probably look into getting better lighting. Keep up the good work!
awswome review !!! THANK YOU!!!
thankss this is really helpful for my assignment
Awesome bro
great great vid!
Great video! You definitely did your homework on this one.
Never really understood the 1x1 convolution.
Thanks Mark. Yeah. This one took a while. Had to do a lot of reading. Plus the set up with the face cam is new. Never done it before because I couldn't get my hands on a decent camera. Still trying to get used to it. I'm glad at least the content (most important part) is useful.
Andrew ng's Convnet course of the deeplearning.ai specialization has a good video on 1x1 convolutions. I think it's in the 2nd or 3rd week of that part of the specialization
hi. thanks for the detailed explanation. i am completely new in this. how is MTCNN if put in the picture?
Love ur videos
Please make a pytorch series
Thanks for watching Samin! pytorch is interesting. Maybe a future video
great
Any link to timeline?
Definitely keep showing your face while presenting. Made this video 50% more dynamic...
You mean the error rate reduced and not the accuracy with overlapping pooling at 07:00
zishan ahmed yup. My bad. Kind of the reason why I wrote it on screen too
i bet this is an indian man
video is already outdated lol deep learning..