English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent breakthroughs in diffusion models, multimodal pretraining, andefficient finetuning have led to an explosion of text-to-image generativemodels. Given human evaluation is expensive and difficult to scale, automatedmethods are critical for evaluating the increasingly large number of newmodels. However, most current automated evaluation metrics like FID orCLIPScore only offer a holistic measure of image quality or image-textalignment, and are unsuited for fine-grained or instance-level analysis. Inthis paper, we introduce GenEval, an object-focused framework to evaluatecompositional image properties such as object co-occurrence, position, count,and color. We show that current object detection models can be leveraged toevaluate text-to-image models on a variety of generation tasks with stronghuman agreement, and that other discriminative vision models can be linked tothis pipeline to further verify properties like object color. We then evaluateseveral open-source text-to-image models and analyze their relative generativecapabilities on our benchmark. We find that recent models demonstratesignificant improvement on these tasks, though they are still lacking incomplex capabilities such as spatial relations and attribute binding. Finally,we demonstrate how GenEval might be used to help discover existing failuremodes, in order to inform development of the next generation of text-to-imagemodels. Our code to run the GenEval framework is publicly available athttps://github.com/djghosh13/geneval.

GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment