Intro | Neural Style Transfer #1

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

Kicking off the neural style transfer series! 🎨

Neural style transfer is in its most basic form about transferring style from a style image onto a content image using neural nets.

In its more complex form you can:
✔️ Transfer style to videos (and additionally use temporal loss)
✔️ Choose whether to keep the color from the content image or take it from the style image
✔️ Use segmentation masks to specify objects which should be styled
and many more cool things - stay tuned for the series!

[Credit] Music:

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
💰 BECOME A PATREON OF THE AI EPIPHANY ❤️

If these videos, GitHub projects, and blogs help you,
consider helping me out by supporting me on Patreon!

Much love! ❤️

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

💡 The AI Epiphany is a channel dedicated to simplifying the field of AI using creative visualizations and in general, a stronger focus on geometrical and visual intuition,
rather than the algebraic and numerical "intuition".

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
👋 CONNECT WITH ME ON SOCIAL

👨‍👩‍👧‍👦 JOIN OUR DISCORD COMMUNITY:

📢 SUBSCRIBE TO MY MONTHLY AI NEWSLETTER:

💻 FOLLOW ME ON GITHUB FOR COOL PROJECTS:

📚 FOLLOW ME ON MEDIUM:
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

#neuralstyletransfer #deeplearning #ai

[makeandbreak]

oh my god, that is awesome, you look like you are animated in that last style. This is awesome. I'm going to watch this series.

[nguyenbaodung]

My new definition of luxuries: binge watch every videos of you

[pleasuretek]

Yo that temporal loss optimization looks awesome. Learning about that is exactly what I want to know. I have experimented with feedforward networks before and the jitter from frame to frame always bugged me so I moved on to other things. Great video series and great work, much love.

[martiancoderone]

Hello dude, your video is so so so so so GREAT! Learned a lot from your tutorial! You a like a golden gem berried in the sand waiting to be discovered! Love your work, just runned your code and it really did awesome job, i'll further do some training on my models and i'll contact u later. GREAT work, again!
dude! May the force be with u!

[pholdway]

Brill concept... Question if A times B gives a certain picture then will B Times A give the identical picture? What about A times B times A times B again? one could do so much with this ! !

[amazing-graceolutomilayo]

Is there any prerequisite you'd advice one to have?

[artificium_]

could the blend technique you used in your latest deep dream script be used for this to reduce flicker? or would the effect be too additive and ruin it. the temporal loss works well for a stationary background but once things get moving it becomes somewhat unstable

[gabrielyashim]

Hello, do you have the code to this video so i can practice and create a replica of this?