Tuesday, June 3, 2025
News PouroverAI
Visit PourOver.AI
No Result
View All Result
  • Home
  • AI Tech
  • Business
  • Blockchain
  • Data Science & ML
  • Cloud & Programming
  • Automation
  • Front-Tech
  • Marketing
  • Home
  • AI Tech
  • Business
  • Blockchain
  • Data Science & ML
  • Cloud & Programming
  • Automation
  • Front-Tech
  • Marketing
News PouroverAI
No Result
View All Result

Text-to-image generation in any style – Google Research Blog

December 15, 2023
in AI Technology
Reading Time: 5 mins read
0 0
A A
0
Share on FacebookShare on Twitter


Posted by: Kihyuk Sohn and Dilip Krishnan, Research Scientists, Google Research

Text-to-image models trained on large volumes of image-text pairs have enabled the creation of rich and diverse images encompassing many genres and themes. Moreover, popular styles such as “anime” or “steampunk”, when added to the input text prompt, may translate to specific visual outputs.

While many efforts have been put into prompt engineering, a wide range of styles are simply hard to describe in text form due to the nuances of color schemes, illumination, and other characteristics. As an example, “watercolor painting” may refer to various styles, and using a text prompt that simply says “watercolor painting style” may either result in one specific style or an unpredictable mix of several. When we refer to “watercolor painting style,” which do we mean?

Instead of specifying the style in natural language, StyleDrop allows the generation of images that are consistent in style by referring to a style reference image*. In this blog we introduce “StyleDrop: Text-to-Image Generation in Any Style”, a tool that allows a significantly higher level of stylized text-to-image synthesis. Instead of seeking text prompts to describe the style, StyleDrop uses one or more style reference images that describe the style for text-to-image generation. By doing so, StyleDrop enables the generation of images in a style consistent with the reference, while effectively circumventing the burden of text prompt engineering. This is done by efficiently fine-tuning the pre-trained text-to-image generation models via adapter tuning on a few style reference images. Moreover, by iteratively fine-tuning the StyleDrop on a set of images it generated, it achieves the style-consistent image generation from text prompts.

Method overview

StyleDrop is a text-to-image generation model that allows generation of images whose visual styles are consistent with the user-provided style reference images. This is achieved by a couple of iterations of parameter-efficient fine-tuning of pre-trained text-to-image generation models. Specifically, we build StyleDrop on Muse, a text-to-image generative vision transformer.

Muse: text-to-image generative vision transformer

Muse is a state-of-the-art text-to-image generation model based on the masked generative image transformer (MaskGIT). Unlike diffusion models, such as Imagen or Stable Diffusion, Muse represents an image as a sequence of discrete tokens and models their distribution using a transformer architecture. Compared to diffusion models, Muse is known to be faster while achieving competitive generation quality.

Parameter-efficient adapter tuning

StyleDrop is built by fine-tuning the pre-trained Muse model on a few style reference images and their corresponding text prompts. There have been many works on parameter-efficient fine-tuning of transformers, including prompt tuning and Low-Rank Adaptation (LoRA) of large language models. Among those, we opt for adapter tuning, which is shown to be effective at fine-tuning a large transformer network for language and image generation tasks in a parameter-efficient manner. For example, it introduces less than one million trainable parameters to fine-tune a Muse model of 3B parameters, and it requires only 1000 training steps to converge.

Parameter-efficient adapter tuning of Muse.

Iterative training with feedback

While StyleDrop is effective at learning styles from a few style reference images, it is still challenging to learn from a single style reference image. This is because the model may not effectively disentangle the content (i.e., what is in the image) and the style (i.e., how it is being presented), leading to reduced text controllability in generation. We address this issue by training a new StyleDrop model on a subset of synthetic images, chosen by the user or by image-text alignment models (e.g., CLIP), whose images are generated by the first round of the StyleDrop model trained on a single image. By training on multiple synthetic image-text aligned images, the model can easily disentangle the style from the content, thus achieving improved image-text alignment.

Iterative training with feedback*. The first round of StyleDrop may result in reduced text controllability, such as a content leakage or concept collapse, due to the difficulty of content-style disentanglement. Iterative training using synthetic images, generated by the previous rounds of StyleDrop models and chosen by human or image-text alignment models, improves the text adherence of stylized text-to-image generation.

Experiments

StyleDrop gallery

We show the effectiveness of StyleDrop by running experiments on 24 distinct style reference images. As shown below, the images generated by StyleDrop are highly consistent in style with each other and with the style reference image, while depicting various contexts, such as a baby penguin, banana, piano, etc. Moreover, the model can render alphabet images with a consistent style.

Stylized text-to-image generation. Style reference images* are on the left inside the yellow box. Text prompts used are:

  • First row: a baby penguin, a banana, a bench.
  • Second row: a butterfly, an F1 race car, a Christmas tree.
  • Third row: a coffee maker, a hat, a moose.
  • Fourth row: a robot, a towel, a wood cabin.

Stylized visual character generation

Style reference images* are on the left inside the yellow box. Text prompts used are: (first row) letter ‘A’, letter ‘B’, letter ‘C’, (second row) letter ‘E’, letter ‘F’, letter ‘G’.

Generating images of my object in my style

Below we show generated images by sampling from two personalized generation distributions, one for an object and another for the style. Images at the top in the blue border are object reference images from the DreamBooth dataset (teapot, vase, dog and cat), and the image on the left at the bottom in the red border is the style reference image*. Images in the purple border (i.e. the four lower right images) are generated from the style image of the specific object.

Quantitative results

For the quantitative evaluation, we synthesize images from a subset of Parti prompts and measure the image-to-image CLIP score for style consistency and image-to-text CLIP score for text consistency. We study non–fine-tuned models of Muse and Imagen. Among fine-tuned models, we make a comparison to DreamBooth on Imagen, state-of-the-art personalized text-to-image method for subjects. We show two versions of StyleDrop, one trained from a single style reference image, and another, “StyleDrop (HF)”, that is trained iteratively using synthetic images with human feedback as described above.

As shown below, StyleDrop (HF) shows significantly improved style consistency score over its non–fine-tuned counterpart (0.694 vs. 0.556), as well as DreamBooth on Imagen (0.694 vs. 0.644). We observe an improved text consistency score with StyleDrop (HF) over StyleDrop (0.322 vs. 0.313). In addition, in a human preference study between DreamBooth on Imagen and StyleDrop on Muse, we found that 86% of the human raters preferred StyleDrop on Muse over DreamBooth on Imagen in terms of consistency to the style reference image.

Conclusion

StyleDrop achieves style consistency at text-to-image generation using a few style reference images. Google’s AI Principles guided our development of Style Drop, and we urge the responsible use of the technology. StyleDrop was adapted to create a custom style model in Vertex AI, and we believe it could be a helpful tool for art directors and graphic designers — who might want to brainstorm or prototype visual assets in their own styles, to improve their productivity and boost their creativity — or businesses that want to generate new media assets that reflect a particular brand. As with other generative AI capabilities, we recommend that practitioners ensure they align with copyrights of any media assets they use.

More results are found on our project website and YouTube video.

Acknowledgements

This research was conducted by Kihyuk Sohn, Nataniel Ruiz, Kimin Lee, Daniel Castro Chin, Irina Blok, Huiwen Chang, Jarred Barber, Lu Jiang, Glenn Entis, Yuanzhen Li, Yuan Hao, Irfan Essa, Michael Rubinstein, and Dilip Krishnan. We thank owners of images used in our experiments (links for attribution) for sharing their valuable assets.

*See image sources



Source link

Tags: BlogGenerationGoogleResearchstyleTexttoImage
Previous Post

Examples of sustainability in business

Next Post

Begin the New Year Clean and Free

Related Posts

How insurance companies can use synthetic data to fight bias
AI Technology

How insurance companies can use synthetic data to fight bias

June 10, 2024
From Low-Level to High-Level Tasks: Scaling Fine-Tuning with the ANDROIDCONTROL Dataset
AI Technology

From Low-Level to High-Level Tasks: Scaling Fine-Tuning with the ANDROIDCONTROL Dataset

June 10, 2024
Decoding Decoder-Only Transformers: Insights from Google DeepMind’s Paper
AI Technology

Decoding Decoder-Only Transformers: Insights from Google DeepMind’s Paper

June 9, 2024
How Game Theory Can Make AI More Reliable
AI Technology

How Game Theory Can Make AI More Reliable

June 9, 2024
Buffer of Thoughts (BoT): A Novel Thought-Augmented Reasoning AI Approach for Enhancing Accuracy, Efficiency, and Robustness of LLMs
AI Technology

Buffer of Thoughts (BoT): A Novel Thought-Augmented Reasoning AI Approach for Enhancing Accuracy, Efficiency, and Robustness of LLMs

June 9, 2024
Deciphering Doubt: Navigating Uncertainty in LLM Responses
AI Technology

Deciphering Doubt: Navigating Uncertainty in LLM Responses

June 9, 2024
Next Post
Begin the New Year Clean and Free

Begin the New Year Clean and Free

Customer service trends winning organizations need to follow

Customer service trends winning organizations need to follow

Rivian Stock Is Risky. Here’s Why I’m Not Buying.

Rivian Stock Is Risky. Here's Why I'm Not Buying.

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

  • Trending
  • Comments
  • Latest
Is C.AI Down? Here Is What To Do Now

Is C.AI Down? Here Is What To Do Now

January 10, 2024
Accenture creates a regulatory document authoring solution using AWS generative AI services

Accenture creates a regulatory document authoring solution using AWS generative AI services

February 6, 2024
Managing PDFs in Node.js with pdf-lib

Managing PDFs in Node.js with pdf-lib

November 16, 2023
23 Plagiarism Facts and Statistics to Analyze Latest Trends

23 Plagiarism Facts and Statistics to Analyze Latest Trends

June 4, 2024
Azul cloud service spots dead code in Java apps

Azul cloud service spots dead code in Java apps

October 7, 2023
The 15 Best Python Courses Online in 2024 [Free + Paid]

The 15 Best Python Courses Online in 2024 [Free + Paid]

April 13, 2024
Can You Guess What Percentage Of Their Wealth The Rich Keep In Cash?

Can You Guess What Percentage Of Their Wealth The Rich Keep In Cash?

June 10, 2024
AI Compared: Which Assistant Is the Best?

AI Compared: Which Assistant Is the Best?

June 10, 2024
How insurance companies can use synthetic data to fight bias

How insurance companies can use synthetic data to fight bias

June 10, 2024
5 SLA metrics you should be monitoring

5 SLA metrics you should be monitoring

June 10, 2024
From Low-Level to High-Level Tasks: Scaling Fine-Tuning with the ANDROIDCONTROL Dataset

From Low-Level to High-Level Tasks: Scaling Fine-Tuning with the ANDROIDCONTROL Dataset

June 10, 2024
UGRO Capital: Targeting to hit milestone of Rs 20,000 cr loan book in 8-10 quarters: Shachindra Nath

UGRO Capital: Targeting to hit milestone of Rs 20,000 cr loan book in 8-10 quarters: Shachindra Nath

June 10, 2024
Facebook Twitter LinkedIn Pinterest RSS
News PouroverAI

The latest news and updates about the AI Technology and Latest Tech Updates around the world... PouroverAI keeps you in the loop.

CATEGORIES

  • AI Technology
  • Automation
  • Blockchain
  • Business
  • Cloud & Programming
  • Data Science & ML
  • Digital Marketing
  • Front-Tech
  • Uncategorized

SITEMAP

  • Disclaimer
  • Privacy Policy
  • DMCA
  • Cookie Privacy Policy
  • Terms and Conditions
  • Contact us

Copyright © 2023 PouroverAI News.
PouroverAI News

No Result
View All Result
  • Home
  • AI Tech
  • Business
  • Blockchain
  • Data Science & ML
  • Cloud & Programming
  • Automation
  • Front-Tech
  • Marketing

Copyright © 2023 PouroverAI News.
PouroverAI News

Welcome Back!

Login to your account below

Forgotten Password? Sign Up

Create New Account!

Fill the forms bellow to register

All fields are required. Log In

Retrieve your password

Please enter your username or email address to reset your password.

Log In