Count objects in image with zero examples using patches like a generated sample with ZSC
Count objects in image with zero examples using patches like a generated sample with ZSC
Zero-shot Object Counting
arXiv paper abstract https://arxiv.org/abs/2303.02001
arXiv PDF paper https://arxiv.org/pdf/2303.02001.pdf
Class-agnostic object counting aims to count object instances of an arbitrary class at test time … Current methods require human-annotated exemplars as inputs which are often unavailable for novel categories
… propose zero-shot object counting (ZSC) … where only the class name is available during test time … does not require human annotators … and can operate automatically.
… propose a method that can accurately identify the optimal patches which can then be used as counting exemplars.
… first construct a class prototype to select the patches that are likely to contain the objects of interest, namely class-relevant patches.
… quantitatively measure how suitable an arbitrary patch is as a counting exemplar … select the most suitable patches as exemplars
… Experimental results on a recent class-agnostic counting dataset, FSC-147, validate the effectiveness of … method …
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
