OpenAI - Jukebox

  • 8 Replies
  • 2296 Views
*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
OpenAI - Jukebox
« on: May 01, 2020, 02:50:13 am »
https://openai.com/blog/jukebox/

They have been quite for so long on the main front page. Never expected that from their past activity!

While AIVA generates comparable music at the moment, it is MIDI only I think, and the voice cloner seen from Two Minute Papers is perfect - why don't they use that technology!? Anyway it's good.

My favorite is the very end of https://jukebox.openai.com/?song=805378117
« Last Edit: May 02, 2020, 04:46:11 am by LOCKSUIT »
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: OpenAI - Jukebox
« Reply #1 on: May 02, 2020, 12:16:37 am »
Oh, this is impressive!
I found proof it works.
There is a multiple completion tries - and each is different and sounds great!
https://jukebox.openai.com/?song=787731448
I assume we can now generate our own related music and extend songs too at parts we want to change within a track.
« Last Edit: May 02, 2020, 04:46:30 am by LOCKSUIT »
Emergent          https://openai.com/blog/

*

yotamarker

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1003
  • battle programmer
    • battle programming
Re: OpenAI - Jukebox
« Reply #2 on: May 03, 2020, 04:09:24 am »
how is this different from vocaloid tho ?



( exept for vocaloid mogging them to oblivion )

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: OpenAI - Jukebox
« Reply #3 on: May 03, 2020, 04:39:18 am »
Well, vocaloid can't generates Katy Perry's voice and all sorts of stuff, fast too. AI does things faster than us we can.
Emergent          https://openai.com/blog/

*

yotamarker

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1003
  • battle programmer
    • battle programming
Re: OpenAI - Jukebox
« Reply #4 on: May 03, 2020, 03:57:29 pm »
Well, vocaloid can't generates Katy Perry's voice and all sorts of stuff, fast too. AI does things faster than us we can.

actually, vocaloid just needs some phonem recordings

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: OpenAI - Jukebox
« Reply #5 on: May 06, 2020, 09:58:50 am »
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: OpenAI - Jukebox
« Reply #6 on: May 06, 2020, 05:57:08 pm »
Watch full video:
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: OpenAI - Jukebox
« Reply #7 on: May 06, 2020, 06:23:47 pm »
See above. Also I found more, and omg so coooooooooooooooool must see must see!

the lyrics in this one = LOL:
https://jukebox.openai.com/?song=787833151

INCREDIBLE
https://jukebox.openai.com/?song=789452305

Wow! That was so refreshing. Whew. I'm in rainbow heaven, um. We made it! Wait not yet...


and omega technoz
https://jukebox.openai.com/?song=787969282

and at 0:50
https://jukebox.openai.com/?song=787734292
« Last Edit: May 06, 2020, 08:35:36 pm by LOCKSUIT »
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: OpenAI - Jukebox
« Reply #8 on: June 05, 2020, 08:10:06 pm »
Facebook made a translator for music in 2018, interesting!


If GPT-2 recognizes the last sentence to predict the next word, and Jukebox does as well (but music), doesn't that mean Jukebox can find the closest translations of the sequence / music? You could take a 5 second long piece of your favorite song and get a lot more closer matches. You could do a lot I imagine (like that video showing faces transition from white to black, Asian to European while standing on their own).
Emergent          https://openai.com/blog/

 


LLaMA2 Meta's chatbot released
by spydaz (AI News )
August 24, 2024, 02:58:36 pm
ollama and llama3
by spydaz (AI News )
August 24, 2024, 02:55:13 pm
AI controlled F-16, for real!
by frankinstien (AI News )
June 15, 2024, 05:40:28 am
Open AI GPT-4o - audio, vision, text combined reasoning
by MikeB (AI News )
May 14, 2024, 05:46:48 am
OpenAI Speech-to-Speech Reasoning Demo
by MikeB (AI News )
March 31, 2024, 01:00:53 pm
Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 am
Google Bard report
by ivan.moony (AI News )
February 14, 2024, 04:42:23 pm
Elon Musk's xAI Grok Chatbot
by MikeB (AI News )
December 11, 2023, 06:26:33 am

Users Online

305 Guests, 0 Users

Most Online Today: 464. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles