paper:2303.02001v2 (arxiv.org)
code:cvlab-stonybrook/zero-shot-counting: CVPR2023 Zero-shot Counting (github.com)
paper with code:Zero-shot Object Counting | Papers With Code
【组会论文2023.06.16】Zero-Shot Object Counting阅读笔记-CSDN博客
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.
我们提出了零样本目标计数(ZSC),这是一种新的设定,其中在测试时只有类别名称是已知的。这样的计数系统不需要人为注释员的参与,可以自动运行。从一个类别名称开始,我们提出了一种方法,可以准确识别出最优的图像块,这些图像块可以用作计数的样本。我们首先构建一个类别原型,以选择可能包含感兴趣目标的图像块,然后训练一个模型定量测量任意图像块作为计数样本的适用性。通过将此模型应用于所有候选图像块,我们可以选择最适合作为计数样本的图像块。
摘要
Class-agnostic object counting aims to count object instances of an arbitrary class at test time. Current methods for this challenging problem require human-annotated exemplars as inputs, which are often unavailable for novel categories, especially for autonomous systems. Thus, 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. Specifically, we first construct a class prototype to select the patches that are likely to contain the objects of interest, namely class-relevant patches. Furthermore, we introduce a model that can 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. Experimental results on a recent class-agnostic counting dataset, FSC-147, validate the effectiveness of our method. Code is available at https://github.com/cvlabstonybrook/zero-shot-counting.
类别无关的目标计数旨在测试时计算任意类别的目标实例。当前针对这一具有挑战性问题的方法需要人为注释的样本作为输入,而这些样本对于新类别尤其是自主系统来说通常不可用。因此,我们提出了零样本目标计数(ZSC),这是一种新的设定,其中在测试时只有类别名称是已知的。这样的计数系统不需要人为注释员的参与,可以自动运行。从一个类别名称开始,我们提出了一种方法,可以准确识别出最优的图像块,这些图像块可以用作计数的样本。具体来说,我们首先构建一个类别原型,以选择可能包含感兴趣目标的图像块,即类别相关的图像块。此外,我们引入了一个模型,可以定量测量任意图像块作为计数样本的适用性。通过将此模型应用于所有候选图像块,我们可以选择最适合作为计数样本的图像块。我们在一个最新的类别无关计数数据集FSC-147上的实验结果验证了我们方法的有效性。