Hacker News new | past | comments | ask | show | jobs | submit login
Video2StyleGAN: Disentangling Local and Global Variations in a Video (arxiv.org)
50 points by lnyan on May 30, 2022 | hide | past | favorite | 5 comments



AI image manipulation and creation is getting really wild. The last 12 months has seen a real explosion in published research. It feels like an exponential curve.


Absolutely. I feel like I never have a good understanding of what is possible anymore because every few months something shows up that blows away all my expectations of what tech can do.

It feels like not so long ago the best we had was the google deep dream stuff that was only good at generating weird nightmare images of dog faces.


Is there anything more useless than these single digit page pdfs full of irreproducible claims?


The paper seems to be about generating videos, where can I view those videos?


"Comments: Video: this https URL" -> https://youtu.be/oUeXFyfdE1A




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: