Different domain images in FSC147 dataset.

<div><p>Few-shot learning techniques have enabled the rapid adaptation of a general AI model to various tasks using limited data. In this study, we focus on class-agnostic low-shot object counting, a challenging problem that aims to achieve accurate object counting with only a few annota...

Full description

Saved in:

Bibliographic Details
Main Author:	Qinghua Yang (606019) (author)
Other Authors:	Bin Liu (5899) (author), Yan Tian (298301) (author), Yangming Shi (21499520) (author), Xinxin Du (470642) (author), Fangyuan He (6455462) (author), Jikun Guo (21499523) (author)
Published:	2025
Subjects:	Sociology Space Science Biological Sciences not elsewhere classified Information Systems not elsewhere classified similar objects within shanghaitech datasets demonstrate scale fusion modules reducing training costs optimize computational efficiency many practical applications learnable shape embedding equally critical factor enhance feature representation encoder applies self relatively little attention integrate adjacent features agnostic counting framework model &# 8217 general ai model shot object counting scale hybrid encoder shot learning techniques hybrid encoder level features attention exclusively agnostic low shot ). xlink "> rapid adaptation level characteristics generalizability compared extensive experiments existing methods enhancing performance efficient low challenging problem annotated samples annotated data achieves real
Tags:	Add Tag No Tags, Be the first to tag this record!

Different domain images in FSC147 dataset.

Similar Items