---
title: Locount
canonical_url: "https://www.modelscope.cn/datasets/OmniData/Locount"
md_url: "https://www.modelscope.cn/datasets/OmniData/Locount.md"
repository: OmniData/Locount
last_updated: 2024-07-14
license: "[Unknown]"
storage_size: "25 GB"
domain:
  - publishDate
  - paperUrl
  - publishUrl
  - displayName
  - taskTypes
  - publisher
tasks:
  - 2020
  - "https://arxiv.org/pdf/2003.08230v3.pdf"
  - "https://isrc.iscas.ac.cn/gitlab/research/locount-dataset"
  - Locount
  - "Object Detection"
  - "State Key Laboratory of Computer Science, Institute of Software Chinese Academy of Sciences"
downloads: 5228
stars: 0
---

# Locount

> Locount - OmniData 在 ModelScope 开源的数据集。displayName: Locount license: Unknown paperUrl: https://arxiv.org/pdf/2003.08230v3.pdf publishDate: "2020" publishUrl: https://isrc.iscas.ac.cn/gitlab/research/locount-dataset publisher: University of Chinese Academy of…

OmniData/Locount 是 ModelScope 魔搭社区上的2020、https://arxiv.org/pdf/2003.08230v3.pdf、https://isrc.iscas.ac.cn/gitlab/research/locount-dataset数据集，涉及 publishDate、paperUrl、publishUrl 领域，存储大小 25 GB，采用 [Unknown] 许可。

- **Repository**: OmniData/Locount
- **License**: [Unknown]
- **Tasks**: 2020, https://arxiv.org/pdf/2003.08230v3.pdf, https://isrc.iscas.ac.cn/gitlab/research/locount-dataset, Locount, Object Detection, State Key Laboratory of Computer Science, Institute of Software Chinese Academy of Sciences
- **Domain**: publishDate, paperUrl, publishUrl, displayName, taskTypes, publisher
- **Storage size**: 25 GB
- **Downloads**: 5228
- **Stars**: 0
- **Last updated**: 2024-07-14

Source: https://www.modelscope.cn/datasets/OmniData/Locount

---

displayName: Locount
license:
- Unknown
paperUrl: https://arxiv.org/pdf/2003.08230v3.pdf
publishDate: "2020"
publishUrl: https://isrc.iscas.ac.cn/gitlab/research/locount-dataset
publisher:
- University of Chinese Academy of Sciences
- ByteDance
- State Key Laboratory of Computer Science, Institute of Software Chinese Academy
  of Sciences
tags:
- Goods
taskTypes:
- Object Counting
- Object Detection

---
# 数据集介绍
  ## 简介
  我们在 28 家不同的商店和公寓中收集了一个大规模的对象定位和计数数据集，其中包含 50,394 张图像，JPEG 图像分辨率为 1920x1080 像素。注释了 140 个类别（包括夹克、鞋子、烤箱等）中的超过 190 万个对象实例。
  ## 引文
  ```
@inproceedings{cai2021rethinking,
  title={Rethinking object detection in retail stores},
  author={Cai, Yuanqiang and Wen, Longyin and Zhang, Libo and Du, Dawei and Wang, Weiqiang},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={35},
  number={2},
  pages={947--954},
  year={2021}
  }
```
  
## Download dataset
:modelscope-code[]{type="git"}