• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Lucas Beyer Calls 2015 DeepDream Experiments Coolest Thing

    AI researcher replies to original creators praising the Inceptionism infinite zoom videos.

    LB
    SD
    AM
    3 Sources, 31d ago, first seen 31d ago

    TLDR

    Lucas Beyer, an AI researcher at Meta formerly at OpenAI and DeepMind, replied to @sedielem, @317070, @avdnoord, @lpigou and @GoogleDeepMind. He stated that the 2015 adaptation of Google's Inceptionism technique using VGG16 to generate infinite zooming psychedelic videos was one of the coolest things back then. Beyer thanked the creators for making it. The reply appears among visible posts referencing those early experiments.

    Combined views

    23.6K

    3 Sources, first seen 31d ago

    Combined views

    23.6K

    3 Sources, first seen 31d ago

    182 likes
    182 likes
    5 comments
    66 saves
    15 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    5 comments
    66 saves
    15 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    @sedielemCool to see people still remember this! Nice reminder of a fun time at the Reservoir Lab back in 2015 with @317070, @avdnoord, @lpigou, right before a bunch of us graduated and joined @GoogleDeepMind. Perhaps unsurprising that a fair few of us work on generative media today😁
    @giffmana@sedielem @317070 @avdnoord @lpigou @GoogleDeepMind It was one of the coolest things back then, thanks for making it!
    @zzznahThanks Lucas! These were crazy 11+ years when things seemed impossible became real! Btw, original DeepDream used InteptionV1 (by @ChrSzegedy), which later became the testbed for early mechanic interpretability. VGG16 powered artistic style transfer - the next big thing in neural art before GANs. Another crazy thing I recently learned about VGG16 is that it's STILL being used as a part of perception loss in latent diffusion autoencoder pretraining!🤯 Heard it's surprisingly difficult to replace, switching to modern dense feature extractors like DINO seems to make almost no difference 🤷

    3 Sources

    @sedielemCool to see people still remember this! Nice reminder of a fun time at the Reservoir Lab back in 2015 with @317070, @avdnoord, @lpigou, right before a bunch of us graduated and joined @GoogleDeepMind. Perhaps unsurprising that a fair few of us work on generative media today😁
    @giffmana@sedielem @317070 @avdnoord @lpigou @GoogleDeepMind It was one of the coolest things back then, thanks for making it!
    @zzznahThanks Lucas! These were crazy 11+ years when things seemed impossible became real! Btw, original DeepDream used InteptionV1 (by @ChrSzegedy), which later became the testbed for early mechanic interpretability. VGG16 powered artistic style transfer - the next big thing in neural art before GANs. Another crazy thing I recently learned about VGG16 is that it's STILL being used as a part of perception loss in latent diffusion autoencoder pretraining!🤯 Heard it's surprisingly difficult to replace, switching to modern dense feature extractors like DINO seems to make almost no difference 🤷