ebook img

Photo Editing With Generative Adversarial Networks (GANs) PDF

25 PagesΒ·2017Β·1.31 MBΒ·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Photo Editing With Generative Adversarial Networks (GANs)

Photo Editing With Generative Adversarial Networks (GANs) GTC, May 2017. Greg Heinrich. PANDARUS: Alas, I think he shall be come approached and the day When little srain would be attain'd into being never fed, And who is but a chain and subjects of his death, I should not sleep. G : WHAT IS A GENERATIVE MODEL? AN In Machine Learning A generative model learns to generate samples that have the same characteristics as the samples in the dataset. Learn from Shakespeare novels: Produce: http://karpathy.github.io/2015/05/21/rnn- PANDARUS: effectiveness/ Alas, I think he shall be come approached and the day When little srain would be attain'd into being never fed, And who is but a chain and subjects of his death, I should not sleep. 22 BASIC REMINDER: BACKPROP πœ•πΈ Calculating iteratively 𝑙 πœ•π‘€ 𝑖𝑗 Chain rule πœ•πΈ Calculation of : Output of each neuron 𝑗 of layer 𝑙 : 𝑙 πœ•π‘§ 𝑗 𝑙 𝑙 𝑙 π‘™βˆ’1 𝑙 β„Ž = πœ‘ 𝑧 = πœ‘ 𝑀 β„Ž + 𝑏 𝑗 𝑗 𝑖𝑗 𝑖 𝑗 πœ•πΈ πœ•πΈ πœ•π‘§π‘™+1 πœ•πΈ πœ•π‘§π‘™+1 πœ•β„Žπ‘™ = π‘˜ = π‘˜ 𝑗 𝑖 𝑙 π‘˜ 𝑙+1 𝑙 π‘˜ 𝑙+1 𝑙 𝑙 πœ•π‘§ πœ•π‘§ πœ•π‘§ πœ•π‘§ πœ•β„Ž πœ•π‘§ 𝑗 π‘˜ 𝑗 π‘˜ 𝑗 𝑗 Gradient of E with respect to each weight: πœ•πΈ 𝑙+1 𝑙 = 𝑀 πœ‘β€² 𝑧 πœ•πΈ πœ•πΈ πœ•π‘§π‘™ πœ•πΈ πœ•π‘§π‘™+1 π‘—π‘˜ 𝑗 𝑗 π‘™βˆ’1 π‘˜ = = β„Ž π‘˜ 𝑙 𝑙 𝑙 𝑙 𝑖 πœ•π‘€ πœ•π‘§ πœ•π‘€ πœ•π‘§ πœ•πΈ 𝑖𝑗 𝑗 𝑖𝑗 𝑗 𝑙 𝑙+1 = πœ‘β€² 𝑧 𝑀 Calculated 𝑗 𝑙+1 π‘—π‘˜ πœ•π‘§ during π‘˜ π‘˜ forward prop Multivariate chain rule 𝑙 Chain πœ•π‘§ 𝑗 only depends on β„Žπ‘™βˆ’1 rule πœ•π‘€π‘™ 𝑖 33 𝑖𝑗 A : PLAYING THE ADVERSARIAL GAME G N Learning on a corpus of images Let’s play a game opposing two agents: - The Generator, a little imp in the computer who paints images. - The Discriminator: you are collectively responsible for playing the Discriminator. The game master (me) randomly picks images from either the corpus or the Generator and shows them to the Discriminator. The goal of the Discriminator is to identify the source of the images: real (from the corpus) or fake (painted by the little imp). The goal of the Generator is to fool the Discriminator. 44 PLAYING THE ADVERSARIAL GAME Is this a veelhoek* from our corpus? Note: you don’t have to * veelhoek is the articulation know what a veelhoek is, of a ubiquitous item in the you will learn through language of a tiny country in examples! Europe that is well known for the inferior quality of its cheese. Yes, this red square is a veelhoek! 55 PLAYING THE ADVERSARIAL GAME Is this a veelhoek from our corpus? No, those squiggly lines aren’t right! 66 PLAYING THE ADVERSARIAL GAME Is this a veelhoek from our corpus? Yes, even though it’s blue and tiny! 77 PLAYING THE ADVERSARIAL GAME Is this a veelhoek from our corpus? No, those rounded corners are a giveaway! 88 PLAYING THE ADVERSARIAL GAME Is this a veelhoek from our corpus? No, but it’s a very good fake! 99 PLAYING THE ADVERSARIAL GAME Is this a veelhoek from our corpus? No, it’s the same fake as before! 1100

Description:
Learning on a corpus of images. Let's play a game opposing two agents: - The Generator, a little imp in the computer who paints images.
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.