Mr DeepFakes Forums

If you cannot log into the forums, please change your password to sync your account: You are not allowed to view links. Register or Login to view. Please contact us if you have any issues. (January 11)

Guests - Sign up to view links and videos on the forums. Please check your spam mail. (February 26, 2018)

  • New and improved dark forum theme!
  • Guests can now comment on videos on the tube.
   
  •  Previous
  • 1
  • ...
  • 37
  • 38
  • 39(current)
  • 40
  • 41
  • ...
  • 44
  • Next 

Shoutbox archive

Go to last read | Mark all read

avatar
@surya i just shot you an email. working on extracting more faces for face A. i'll just add them to the Face A folder for you. thanks for the help!
avatar
@surya Can you share more face sets and models that you have trained like the deepika one?
avatar
@wifepics by the way move to LINUX , particulary the liux mint 19 tara distro, initially the learning curve is little steep but eventualy i got the hang of it, and linux is built for AI/Deeplearning
avatar
@wifepics sorry man i had promosed you i would do it ,got caught in something else,please do re- forward me both the face set and would train it for the original model most probalby would finish by sunday
avatar
anyone with a GPU and a heart want to download the Python app I use and train a model for me? I think it should just take an overnight. I've started retraining with my CPU. so if you download the app, i'll send you Face A, Face B, and the model
avatar
so i hate windows auto update. started updating without asking, while i was training. andt it totally just corrupted a model i had been training for a week (yay CPU training)
avatar
@surya are you sure? They look like different videos to me.
avatar
@dpfks you need to delete/remove the one title Shruti Haasan wrongly duplicated with Sonam Kapoor blacked
avatar
i just tried on a two minute clip 50k iteration, weired result, would further train it for another 50 and see,by the way i got GTX1080TI/GTX1070, 8GB ram
avatar
@surya I haven't tried GAN because I think it is too resource intensive and my GPU won't be able to handle it
avatar
i think you have to have a pretty sophisticade system that essentially replicates a neural network. maybe several GPUs in one system.
avatar
Hi folks,have any one of you ever tried the GAN model ? appreciate the feedback
avatar
Note: We are still experiencing some issues with the server. Looks like we have been attacked consistently in the past couple weeks making the website slow. We are still working behind the scenes to fix this.
avatar
@Cerciusx clear your cache In settings and reload the page!
avatar
@chinpope sorry for the typos, the chatbox has a weird wraparound bug on my browser
avatar
@chinpope I alwas wondered about that. He made suc an effort to show before and after photos for no reason, and to me it never looked like he did a good job, ow I'm no genius at deepfaking but thats how I saw it.
avatar
@surya I've been holding Kirti sanon Pornstar look-alike for a week now and didn't put into the list because she's doesn't look enough to be as a doppelganger of her, but I see you didn't find the best match of her, try any of these and let me know .. Camila saint, Francesca Lombardi, Serena Wood
avatar
@dpfks Still UnObtrusiveBot's fakes with that many likes for each of his fakes are extreme outliers to the typical engagement on the site, it wouldn't surprise me that he had other ghost accounts for that many likes. It was obvious he was gaming likes on dpfak when he used to post there.
avatar
There was currently a bug with he voting system, it was supposed to be one per IP, so now I changed it to one per account.
avatar
@chinpope I just changed the voting to one per account. We'll likely see a drastic change in number of votes. I hightly double people would "bot" votes on this site

  •  Previous
  • 1
  • ...
  • 37
  • 38
  • 39(current)
  • 40
  • 41
  • ...
  • 44
  • Next