2025, Vol. 10, No. 3. - go to content...

Permanent address of this page - https://kostumologiya.ru/en/05ivkl325.html

Метаданные этой статьи так же доступны на русском языке

Full article in PDF format (file size: 627.5 KB)

For citation:

Korobtseva N.A., Nagay S.S., Pavlinov A.M., Pchelin G.M. Methodology for creating an animated fashion collection using the Sora generative model. Journal of Clothing Science. 2025; 10(3). Available at: https://kostumologiya.ru/PDF/05IVKL325.pdf (in Russian).

Methodology for creating an animated fashion collection using the Sora generative model

Korobtseva Nadezhda Alekseevna
Russian State University named A.N. Kosygin (Technologies. Design. Art), Moscow, Russia
E-mail: rrr-home@yandex.ru
ORCID: https://orcid.org/0000-0001-9895-6761

Nagay Svetlana Samdorevna
Russian State University named A.N. Kosygin (Technologies. Design. Art), Moscow, Russia
E-mail: lana.nagaina@yandex.ru
ORCID: https://orcid.org/0009-0005-5173-6688

Pavlinov Aleksandr Mikhailovich
Russian State University named A.N. Kosygin (Technologies. Design. Art), Moscow, Russia
E-mail: aleksanderpavlinov@yandex.ru
ORCID: https://orcid.org/0009-0005-7717-9891

Pchelin Grigory Mikhailovich
Russian State University named A.N. Kosygin (Technologies. Design. Art), Moscow, Russia
E-mail: 237285@stud.rguk.ru
ORCID: https://orcid.org/0009-0007-9247-7381

Abstract. This article presents a methodology for creating animated fashion collections using the generative video model Sora. The relevance of the research is driven by the rapid development of artificial intelligence technologies that are significantly transforming visual practices in fashion and design. Contemporary designers are increasingly turning to digital formats, where visual storytelling becomes a central element of presentation. The proposed methodology enables the creation of video-based presentations of fashion collections without requiring skills in 3D graphics or programming. The process is based on working with textual descriptions of fashion looks, which are interpreted by the neural model and transformed into animated scenes. The methodology includes several stages: analyzing visual references, composing textual prompts, interacting with the model, and final video editing. It was tested on a real digital fashion collection and demonstrated its applicability in both professional and educational contexts. Particular attention is given to limitations, such as the model’s subjectivity in interpreting input texts and its reliance on precise verbal formulation. The results confirm the potential of generative models as tools for digital fashion creation and expand the capabilities of designers in the context of post-physical garment presentation.

The developed algorithm demonstrated how a digital fashion collection can be quickly transformed into a visually expressive animated show without resorting to labor-intensive 3D packages. Translating static visual images into text descriptions simplifies the iterative process and allows focusing on artistic aspects rather than technical routine.

Keywords: generative design; digital fashion; video presentation; Sora; visual image; text-to-video; methodology

Download article in PDF format