1:05 What is Discriminator Overfitting? 2:10 “Leaking” Augmentations in GANs 3:50 Invertible Transformations that don’t leak 5:57 Consistency Regularization 7:32 Discriminator Augmentation / Goggles 8:46 Adaptive Control Scheme with Overfitting Heuristics 10:11 Results 13:10 Comparison with Training from Scratch and Transfer Learning in GANs with Freeze-Discriminator
@MrRobotoDancer4 жыл бұрын
Henry i love your videos, you help me a lot with my thesis project. If i had more time i would definitely add this paper to the project. Keep it up with the great work, also if you add this timestamps in the description, youtube automatically adds cuts in the video timeline
@TheThirdLieberkind4 жыл бұрын
@@MrRobotoDancer I was about to say the same - finally a great new feature from youtube!
@ricardokylo72493 жыл бұрын
A tip: you can watch movies on kaldroStream. Me and my gf have been using it for watching loads of movies during the lockdown.
@anthonythomas4893 жыл бұрын
@Ricardo Kylo Yea, been watching on kaldroStream for years myself :D
@lewisbrantley33083 жыл бұрын
@Ricardo Kylo Yea, have been using kaldroStream for since december myself :D
@maverick30694 жыл бұрын
I love these videos! Thanks for bringing the videos on latest research in the field. Keep'em coming!
@Schematical4 жыл бұрын
Interesting. Going to have to try this in the next line of expirements on my channel. Thanks for the info!
@connor-shorten4 жыл бұрын
Awesome! Please let me know if you write a blog post about your experiments!
@SeucheAchat4 жыл бұрын
Your channel is sooo important for me atm :)
@catherinele84172 жыл бұрын
Thanks a lot for this video . Your video helps a lot !
@TimScarfe4 жыл бұрын
Awesome work!
@mehermanoj454 жыл бұрын
Thanks you. 🙂
@connor-shorten4 жыл бұрын
Thanks for watching!
@tpocomp4 жыл бұрын
Thank You, so good video
@rajshreesrivastava76502 жыл бұрын
Hello sir, Nice explanation sir Can u plZ tell how we can use gan for data augmentation and deep learning alexnet /resnet-50/ vgg 16 for classification
@kedirwaritu2156 Жыл бұрын
have you get the answer for this question how you do it please replay me?
@nguyenanhnguyen76583 жыл бұрын
Only work with very focused portrai-alike and face mostly dataset. Any other is crap result. Not usable.