Toward Verifiable and Reproducible Human Evaluation for Text-to-Image Generation
This addresses the need for verifiable and reproducible evaluation methods for researchers and practitioners in text-to-image generation, though it is incremental as it builds on existing evaluation practices.
The paper tackles the problem of unreliable human evaluation in text-to-image generation by proposing a standardized protocol, showing that automatic measures like FID are incompatible with human perception in a pilot study.
Human evaluation is critical for validating the performance of text-to-image generative models, as this highly cognitive process requires deep comprehension of text and images. However, our survey of 37 recent papers reveals that many works rely solely on automatic measures (e.g., FID) or perform poorly described human evaluations that are not reliable or repeatable. This paper proposes a standardized and well-defined human evaluation protocol to facilitate verifiable and reproducible human evaluation in future works. In our pilot data collection, we experimentally show that the current automatic measures are incompatible with human perception in evaluating the performance of the text-to-image generation results. Furthermore, we provide insights for designing human evaluation experiments reliably and conclusively. Finally, we make several resources publicly available to the community to facilitate easy and fast implementations.