Counting without human annotated exemplars.
This repository contains the implementation of the paper Zero-Shot Object Counting. We propose zero-shot object counting (ZSC), a new setting where only the class name is available during test time. Such a counting system does not require human annotators in the loop and can operate automatically. Starting from a class name, we propose a method that can accurately identify the optimal patches which can then be used as counting exemplars. We first construct a class prototype to select the patches that are likely to contain the objects of interest and then we train a model to quantitatively measure how suitable an arbitrary patch is as a counting exemplar. By applying this model to all the candidate patches, we can select the most suitable patches as exemplars for counting.
Please cite our CVPR 2022 paper:
@InProceedings{Xu_2023_CVPR,
author = {Xu, Jingyi and Le, Hieu and Nguyen, Vu and Ranjan, Viresh and Samaras, Dimitris},
title = {Zero-Shot Object Counting},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2023},
pages = {15548-15557}
}
This repo heavily based on BMNet. Thanks for the great work.