Logo Logo
deutsch
directory schraeg
Knowledge
Codecs
Hardware
Camcorders· Cinema-Kamera· Computers· Drohnen· GPU· Kamera-Zubehör· Video-DSLR· accessories
Software
Compositing· Color correction· DV Editing
DV-Movies

HowTo
Shooting· Sound
Misc
Forschung· KI· Reviews· Streaming
/// News
New Nvidia AI generates images 30x faster than Stable Diffusion

New Nvidia AI generates images 30x faster than Stable Diffusion

[16:10 Wed,8.March 2023   by ]    

A team of researchers from Nvidia and the University of Tübingen has presented a new (old) method for generating images via AI, which has two major advantages over the previously used method. The new method, christened StyleGAN-T, is based quite classically on GANs (Generative Adversarial Network), which has long been used by Nvidia in the form of StyleGAN for several interesting applications, such as already in 2018 for the generation of arbitrary human faces or the artificial aging of faces.


StyleGAN-T-Morph


This method, thanks to the new research, can now also generate images by prompt, much like current image AIs, such as DALL-E2, Midjourney, or the open source and therefore most widely used Stable Diffusion. These all rely on the latent diffusion model to generate images from text. In contrast, the new StyleGAN-T method, which has been christened StyleGAN-T, has two particular positive features: it is significantly faster and can morph better between the results of different prompts.

How fast StyleGAN-T is (and how well it can switch from almost smoothly one image to another) is well demonstrated by the following video, in which each 512 x 512 pixel frame was generated in 0.1 seconds - i.e. almost in real time - on a Nvidia A100 Tensor Core GPU:



The difference in the speed of image generation also becomes clear in the following overview, a comparison with other image AIs such as Stable Diffusion and Google&s Imagen. Here, the new method is 37x faster than Stable Diffusion and over 90x faster than Imagen.

NVIDIA-KI-Benchmarks


On weaker GPUs than the highly specialized Nvidisa NVIDIA A100, image generation runs slower than 0.1 seconds per image, of course, but the relative speed compared to images generated by diffusion should remain about the same.

The second advantage of generating images via GANs is their typical ability to explore latent space, i.e., interpolate between different results and thus visually morph between different results at the same or even different text prompts to make them blend seamlessly.

This general capability of (image) GANs is demonstrated quite nicely in the following clip from 2 Minute Paper:



Bild zur Newsmeldung:
StyleGAN-T-Architektur

Link more infos at bei sites.google.com

deutsche Version dieser Seite: Neue Nvidia-KI generiert Bilder 30x schneller als Stable Diffusion

  

  Vorige News lesen Nächste News lesen 
bildGünstiger: S35-Versionen der RED V-RAPTOR (XL) 8K Kameras offiziell vorgestellt bildKIs imitieren Stimmen perfekt - Vorsicht vor anrollenden Betrugswellen


related news:1E0Inspiring AI image variations at the click of a mouse with Stable Diffusion Reimagine / unCLIP 31.March 2023
Insta360 Flow - smartphone gimbal with AI tracking introduced 29.March 2023
Runway Gen-1 available- expensive AI video generator to get you started 28.March 2023
This is what the AI-generated winning films of the Runway AI Film Festival look like 27.March 2023
How to produce a SF clip in just 3h using AI tools 24.March 2023
Microsoft with generative AI models for everyone in the browser - Bing Image Creator 23.March 2023
Adobe announces Firefly: Generative AI models now directly from Adobe 21.March 2023
alle Newsmeldungen zum Thema KI
1E0NVIDIA secretly adds even more parallel video encoding to its GPUs 30.March 2023
VideoFusion: First open source video AI is here - and also runs on the home PC 20.March 2023
Apple: New Macs before launch - is the new Mac Pro with M2 Ultra finally coming? 10.March 2023
Stream more effectively: FFmpeg 6.0 brings among other things acceleration of AV1 via GPU 1.March 2023
Free AI tool NVIDIA Canvas: design photorealistic dream landscapes yourself 15.February 2023
More beautiful time-lapse shots thanks to new NVIDIA AI 11.February 2023
Have fun with NVIDIA's new eye contact effect: Robert De Niro looks you in the eye 27.January 2023
alle Newsmeldungen zum Thema Computers
1E0This is what the AI-generated winning films of the Runway AI Film Festival look like 27.March 2023
Sennheiser Profile USB Microphone - easy to use for podcasts & co 25.March 2023
How to produce a SF clip in just 3h using AI tools 24.March 2023
This could be expensive: Beware of photo wallpaper during film shooting or photo shoot 5.March 2023
Ruben Östlund: Why the pitch is the crucial step in film development 4.March 2023
Home-made virtual film studio: a little test 22.February 2023
Cornerstones of a new film funding: more creativity, less bureaucracy? 19.February 2023
alle Newsmeldungen zum Thema Film


[nach oben]
















Archiv Newsmeldungen

2023

March - February - January

2022
December - November - October - September - August - July - June - May - April - March - February - January

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000






































deutsche Version dieser Seite: Neue Nvidia-KI generiert Bilder 30x schneller als Stable Diffusion



last update : 31.März 2023 - 18:02 - slashCAM is a project by channelunit GmbH- mail : slashcam@--antispam:7465--slashcam.de - deutsche Version