@inproceedings{vladimir-etal-2024-kandinsky,
title = "Kandinsky 3: Text-to-Image Synthesis for Multifunctional Generative Framework",
author = "Vladimir, Arkhipkin and
Vasilev, Viacheslav and
Filatov, Andrei and
Pavlov, Igor and
Agafonova, Julia and
Gerasimenko, Nikolai and
Averchenkova, Anna and
Mironova, Evelina and
Anton, Bukashkin and
Kulikov, Konstantin and
Kuznetsov, Andrey and
Dimitrov, Denis",
editor = "Hernandez Farias, Delia Irazu and
Hope, Tom and
Li, Manling",
booktitle = "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.emnlp-demo.48",
pages = "475--485",
abstract = "Text-to-image (T2I) diffusion models are popular for introducing image manipulation methods, such as editing, image fusion, inpainting, etc. At the same time, image-to-video (I2V) and text-to-video (T2V) models are also built on top of T2I models. We present Kandinsky 3, a novel T2I model based on latent diffusion, achieving a high level of quality and photorealism. The key feature of the new architecture is the simplicity and efficiency of its adaptation for many types of generation tasks. We extend the base T2I model for various applications and create a multifunctional generation system that includes text-guided inpainting/outpainting, image fusion, text-image fusion, image variations generation, I2V and T2V generation. We also present a distilled version of the T2I model, evaluating inference in 4 steps of the reverse process without reducing image quality and 3 times faster than the base model. We deployed a user-friendly demo system in which all the features can be tested in the public domain. Additionally, we released the source code and checkpoints for the Kandinsky 3 and extended models. Human evaluations show that Kandinsky 3 demonstrates one of the highest quality scores among open source generation systems.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="vladimir-etal-2024-kandinsky">
<titleInfo>
<title>Kandinsky 3: Text-to-Image Synthesis for Multifunctional Generative Framework</title>
</titleInfo>
<name type="personal">
<namePart type="given">Arkhipkin</namePart>
<namePart type="family">Vladimir</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viacheslav</namePart>
<namePart type="family">Vasilev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrei</namePart>
<namePart type="family">Filatov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Igor</namePart>
<namePart type="family">Pavlov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Julia</namePart>
<namePart type="family">Agafonova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nikolai</namePart>
<namePart type="family">Gerasimenko</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Averchenkova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Evelina</namePart>
<namePart type="family">Mironova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bukashkin</namePart>
<namePart type="family">Anton</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Konstantin</namePart>
<namePart type="family">Kulikov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andrey</namePart>
<namePart type="family">Kuznetsov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Denis</namePart>
<namePart type="family">Dimitrov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: System Demonstrations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Delia</namePart>
<namePart type="given">Irazu</namePart>
<namePart type="family">Hernandez Farias</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tom</namePart>
<namePart type="family">Hope</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Manling</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Text-to-image (T2I) diffusion models are popular for introducing image manipulation methods, such as editing, image fusion, inpainting, etc. At the same time, image-to-video (I2V) and text-to-video (T2V) models are also built on top of T2I models. We present Kandinsky 3, a novel T2I model based on latent diffusion, achieving a high level of quality and photorealism. The key feature of the new architecture is the simplicity and efficiency of its adaptation for many types of generation tasks. We extend the base T2I model for various applications and create a multifunctional generation system that includes text-guided inpainting/outpainting, image fusion, text-image fusion, image variations generation, I2V and T2V generation. We also present a distilled version of the T2I model, evaluating inference in 4 steps of the reverse process without reducing image quality and 3 times faster than the base model. We deployed a user-friendly demo system in which all the features can be tested in the public domain. Additionally, we released the source code and checkpoints for the Kandinsky 3 and extended models. Human evaluations show that Kandinsky 3 demonstrates one of the highest quality scores among open source generation systems.</abstract>
<identifier type="citekey">vladimir-etal-2024-kandinsky</identifier>
<location>
<url>https://aclanthology.org/2024.emnlp-demo.48</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>475</start>
<end>485</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Kandinsky 3: Text-to-Image Synthesis for Multifunctional Generative Framework
%A Vladimir, Arkhipkin
%A Vasilev, Viacheslav
%A Filatov, Andrei
%A Pavlov, Igor
%A Agafonova, Julia
%A Gerasimenko, Nikolai
%A Averchenkova, Anna
%A Mironova, Evelina
%A Anton, Bukashkin
%A Kulikov, Konstantin
%A Kuznetsov, Andrey
%A Dimitrov, Denis
%Y Hernandez Farias, Delia Irazu
%Y Hope, Tom
%Y Li, Manling
%S Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F vladimir-etal-2024-kandinsky
%X Text-to-image (T2I) diffusion models are popular for introducing image manipulation methods, such as editing, image fusion, inpainting, etc. At the same time, image-to-video (I2V) and text-to-video (T2V) models are also built on top of T2I models. We present Kandinsky 3, a novel T2I model based on latent diffusion, achieving a high level of quality and photorealism. The key feature of the new architecture is the simplicity and efficiency of its adaptation for many types of generation tasks. We extend the base T2I model for various applications and create a multifunctional generation system that includes text-guided inpainting/outpainting, image fusion, text-image fusion, image variations generation, I2V and T2V generation. We also present a distilled version of the T2I model, evaluating inference in 4 steps of the reverse process without reducing image quality and 3 times faster than the base model. We deployed a user-friendly demo system in which all the features can be tested in the public domain. Additionally, we released the source code and checkpoints for the Kandinsky 3 and extended models. Human evaluations show that Kandinsky 3 demonstrates one of the highest quality scores among open source generation systems.
%U https://aclanthology.org/2024.emnlp-demo.48
%P 475-485
Markdown (Informal)
[Kandinsky 3: Text-to-Image Synthesis for Multifunctional Generative Framework](https://aclanthology.org/2024.emnlp-demo.48) (Vladimir et al., EMNLP 2024)
ACL
- Arkhipkin Vladimir, Viacheslav Vasilev, Andrei Filatov, Igor Pavlov, Julia Agafonova, Nikolai Gerasimenko, Anna Averchenkova, Evelina Mironova, Bukashkin Anton, Konstantin Kulikov, Andrey Kuznetsov, and Denis Dimitrov. 2024. Kandinsky 3: Text-to-Image Synthesis for Multifunctional Generative Framework. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 475–485, Miami, Florida, USA. Association for Computational Linguistics.