README.md

---
license: cc
dataset_info:
  features:
  - name: image_id
    dtype: int64
  - name: image
    dtype: image
  - name: embedding
    list: float32
    length: 512
  splits:
  - name: train
    num_bytes: 232472562
    num_examples: 696
  download_size: 233200478
  dataset_size: 232472562
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
task_categories:
- feature-extraction
- zero-shot-classification
- image-classification
- image-to-image
tags:
- biology
size_categories:
- n<1K
source for images:
- Francesco/insects-mytwu
---
## Introduction
This dataset provides CLIP embeddings for images of insects, enabling similarity search and content-based retrieval.

## Source
All images were sourced from [Francesco/insects-mytwu](https://huggingface.co/datasets/Francesco/insects-mytwu).

## Code
If you wish to create a similar dataset then feel free to use the code provided here:
https://github.com/rishik18/CLIP_embedding_based_retrieval

## Sample Usage
```python
from datasets import load_dataset

ds_new = load_dataset("hkanade/insect_image_retrieval")
ds_new["train"][0]["image"]
```
## Examples of image retrieval

![image/png](https://cdn-uploads.huggingface.co/production/uploads/660db6274a2ced4b331eda71/w8KkC2tnsetvTdZdw1e1e.png)
![image/png](https://cdn-uploads.huggingface.co/production/uploads/660db6274a2ced4b331eda71/qacAA7mP590hBQraGa116.png)
![image/png](https://cdn-uploads.huggingface.co/production/uploads/660db6274a2ced4b331eda71/VN4a2mQxLVYx37eSYgTDK.png)
![image/png](https://cdn-uploads.huggingface.co/production/uploads/660db6274a2ced4b331eda71/jY8-R71OWeIvLS0onpKYW.png)

## Citation

If you use this dataset, please cite:

> Kanade, H. (2025). *[insect_image_retrieval](https://huggingface.co/datasets/hkanade/insect_image_retrieval)* [Data set]. Hugging Face. Accessed 2025-08-25.

**BibTeX**
```bibtex
@misc{kanade2025insect_image_retrieval,
  author       = {Hrishikesh Kanade},
  title        = {insect_image_retrieval},
  year         = {2025},
  howpublished = {\url{https://huggingface.co/datasets/hkanade/insect_image_retrieval}},
  note         = {Hugging Face dataset; accessed 2025-08-25}
}