README.md

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

# Intro
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.

## Demo
<https://huggingface.co/spaces/Genius-Society/insecta>

## Usage
```python
from modelscope import snapshot_download
model_dir = snapshot_download("Genius-Society/insecta")
```

## Mirror
<https://www.modelscope.cn/models/Genius-Society/insecta>

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