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

With the availability of powerful text-to-image diffusion models, recentworks have explored the use of synthetic data to improve image classificationperformances. These works show that it can effectively augment or even replacereal data. In this work, we investigate how synthetic data can benefit few-shotclass-agnostic counting. This requires to generate images that correspond to agiven input number of objects. However, text-to-image models struggle to graspthe notion of count. We propose to rely on a double conditioning of StableDiffusion with both a prompt and a density map in order to augment a trainingdataset for few-shot counting. Due to the small dataset size, the fine-tunedmodel tends to generate images close to the training images. We propose toenhance the diversity of synthesized images by exchanging captions betweenimages thus creating unseen configurations of object types and spatial layout.Our experiments show that our diversified generation strategy significantlyimproves the counting accuracy of two recent and performing few-shot countingmodels on FSC147 and CARPK.

Semantic Generative Augmentations for Few-Shot Counting