Keeyon Ebrahimi walks through how neural networks actually work, and what we did before neural networks. A convolutional neural network is a series of layers which each takes an input and produces an output. Keeyon describes the convolutional layer, the pooling layer, and the fully-connected layer.
Keeyon works for Clarifai, a company that uses neural networks to support its computer vision API.
Latest comments (11)
There are the really like this post looking the amazing way to free spades games it is the great blog me most of the people have to seen the update.
You're explanation is very elaborated I really enjoyed watching. Thank you.
greate job! I've met with those concepts during university times, but well no one back then could explain it so smooth and so quick. Thank you!
That fact that I was able to grok the concepts without you having to draw or demo, says alot for how well you articulate the ideas. Nice work Keeyon!
One small nit, the CC was a little off at times and could do with a quick edit pass.
Andrew
We need more stuff like this
Pretty interesting! Good job guys, I wanna see more about this technology :)
This is the best explanation I've ever seen about neural networks and the convolutional layer! Thank you!
Agreed. Great, relatable explanation of these complicated but useful models.
Keeyon! Remember when we used to have "Hacking Class" at the Ivory's?? Miss you man!
Hahaha I do remember the Hacking Classes at the Ivory's!!! This is just an extension of that :). Miss you too man, I hope you're doing well!
A lot clicked for me at around 2:30 with the description of convolutions.