---
license: mit
language:
- en
- zh
- la
metrics:
- accuracy
tags:
- biology
---

# 简介 Intro
这是昆虫识别领域的一种目标检测模型，其支持 2037 个不同类别的昆虫，涵盖了可能是目、科、属或种等多个分类级别。该模型旨在解决生物识别领域中的广泛而具有挑战性的问题，对昆虫和部分节肢动物的多样性和复杂性进行了高效处理。核心目标是实现对大规模昆虫图像数据集的准确检测，以确保每个类别都能得到精准的分类和目标检测。这一研究为生态学、农业保护和生物多样性研究等领域提供了一种先进的技术工具。其创新之处在于模型的大规模类别支持，使其在面对未知昆虫或其他节肢动物的实际应用场景中表现出卓越的性能。

This is an object detection model in the realm of insect identification, supporting a diverse range of 2037 insect categories, spanning multiple taxonomic ranks including order, family, genus, and species. Engineered to tackle extensive and challenging issues within biological recognition, the model adeptly manages the diversity and intricacy of insects and select arthropods. The primary objective is the accurate detection within vast insect imagery datasets, ensuring each category receives precise classification and detection. This study introduces a sophisticated technological instrument for domains such as ecological studies, agricultural conservation, and biodiversity research. The innovation of this model is its extensive category support, which endows it with superior performance when encountering unknown insects or other arthropods in real-world applications.

## 在线体验 Demo
<https://www.modelscope.cn/studios/MuGemSt/insecta>

## 使用 Usage
```python
from modelscope import snapshot_download
model_dir = snapshot_download("MuGemSt/insecta")
```

## 维护 Maintenance
```bash
git clone git@hf.co:MuGemSt/insecta
cd insecta
```

## 镜像 Mirror
<https://www.modelscope.cn/models/MuGemSt/insecta>

## 参考引用 Reference
[1] <https://github.com/quarrying/quarrying-insect-id>