A Comprehensive Evaluation of 26 State-of-the-Art Text-to-Image Models

cover
12 Oct 2024

Authors:

(1) Tony Lee, Stanford with Equal contribution;

(2) Michihiro Yasunaga, Stanford with Equal contribution;

(3) Chenlin Meng, Stanford with Equal contribution;

(4) Yifan Mai, Stanford;

(5) Joon Sung Park, Stanford;

(6) Agrim Gupta, Stanford;

(7) Yunzhi Zhang, Stanford;

(8) Deepak Narayanan, Microsoft;

(9) Hannah Benita Teufel, Aleph Alpha;

(10) Marco Bellagente, Aleph Alpha;

(11) Minguk Kang, POSTECH;

(12) Taesung Park, Adobe;

(13) Jure Leskovec, Stanford;

(14) Jun-Yan Zhu, CMU;

(15) Li Fei-Fei, Stanford;

(16) Jiajun Wu, Stanford;

(17) Stefano Ermon, Stanford;

(18) Percy Liang, Stanford.

Abstract and 1 Introduction

2 Core framework

3 Aspects

4 Scenarios

5 Metrics

6 Models

7 Experiments and results

8 Related work

9 Conclusion

10 Limitations

Author contributions, Acknowledgments and References

A Datasheet

B Scenario details

C Metric details

D Model details

E Human evaluation procedure

6 Models

We evaluate 26 recent text-to-image models, encompassing various types (e.g., diffusion, autoregressive, GAN), sizes (ranging from 0.4B to 13B parameters), organizations, and accessibility (open or closed). Table 4 presents an overview of the models and their corresponding properties. In our evaluation, we employ the default inference configurations provided in the respective model’s API, GitHub, or Hugging Face repositories.

This paper is available on arxiv under CC BY 4.0 DEED license.