US 12,254,170 B2
Utilizing a transformer-based generative language model to generate digital design document variations
Oliver Brdiczka, San Jose, CA (US); Alexandru Vasile Costin, Monte Sereno, CA (US); Ionut Mironica, Bucharest (RO); and Vlad-Constantin Lungu-Stan, Buzau (RO)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on May 8, 2023, as Appl. No. 18/313,529.
Application 18/313,529 is a continuation of application No. 17/652,044, filed on Feb. 22, 2022, granted, now 11,644,961.
Prior Publication US 2023/0305690 A1, Sep. 28, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 30/02 (2023.01); G06F 3/04845 (2022.01); G06F 40/106 (2020.01); G06F 40/284 (2020.01); G06T 7/60 (2017.01)
CPC G06F 3/04845 (2013.01) [G06F 40/106 (2020.01); G06F 40/284 (2020.01); G06T 7/60 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
receiving a digital design document generation query via a graphical user interface;
generating, utilizing a language abstraction model, a design language representation from the digital design document generation query by transforming the digital design document generation query into one or more tokens;
generating, utilizing a generative language model comprising a transformer neural network, an additional design language representation from the design language representation; and
generating, utilizing a decoder, a digital design document from the additional design language representation.