Trained model and additional material: https://mega.nz/#!CCRQQYjL!2ZKP39e40Bw5uDegAXeaBKO12dggtgu6ZRB2CX4XRck
References:
[1] https://www.gwern.net/Danbooru2017
[2] https://github.com/joshua-stone/DerPyBooru
[3] https://pair-code.github.io/saliency/
[4] https://www.reddit.com/r/MachineLearning/comments/7sn4p6/
[5] http://norman-ai.mit.edu/
Seems like user “relgukxilef” had similar plans:
Multi-Layer Perception –> Multi-Layer Perceptron
That was fun
Multi-Layer Perception
Do you know you are all my very best algorithms?
I used to wonder what machine learning could be
Until it programmed itself for me
Neural networks
Deep learning
Fuzzy logic
Image recognition
Pattern matching
It’s an easy feat
And AI makes it all complete
Edited
Multi-Layer Perception Multi-Layer Perception
Ah Ah Ah Ah
Multi-Layer Perception : Fisher Information Matrix
And FIM is Fisher Information Matrix, which makes the space of statistical models into a Riemannian Manifold and allows you to do differential geometric things to it, such as Natural Gradient Descent.
I’m glad I’m not the only one to notice that
Multi-Layer Perception is a type of Neural Network, an algorithm mentioned in this image, but with much more than 1 layers of the middle layer between input and output. (It’s alright if you don’t get this, unless you are in computer science field of study taking this path, this is a tough thing to understand)
Multi-Layer Perception: Friendship is Magic
Cut and paste the ostrich picture’s noise and tada, the school bus is an ostrich
Edited