# Image-Text Retrieval Embdding with CLIP

*author: David Wang*


<br />



## Description

This operator extracts features for image or text with [CLIP](https://arxiv.org/abs/2108.02927) which can genearte the embedding for text and image by jointly training an image encoder and text encoder to maximize the cosine similarity. This operator is an adaptation from [openai/CLIP](https://github.com/openai/CLIP).


<br />


## Code Example

Load an image from path './teddy.jpg' to generate an image embedding. 
Read the text 'a dog' to generate an text embedding. 
 *Write the pipeline in simplified style*:

```python
import towhee

towhee.glob('./dog.jpg') \
      .image_decode.cv2() \
      .towhee.clip(name='ViT-B/32', modality='image') \
      .show()

towhee.dc(["a dog"]) \
      .image_decode.cv2() \
      .towhee.clip(name='ViT-B/32', modality='text') \
      .show()
```
<img src="https://towhee.io/towhee/clip/raw/branch/main/vec1.png" alt="result1" style="height:20px;"/>
<img src="https://towhee.io/towhee/clip/raw/branch/main/vec2.png" alt="result2" style="height:20px;"/>

*Write a same pipeline with explicit inputs/outputs name specifications:*

```python
import towhee

towhee.glob['path']('./dog.jpg') \
      .image_decode.cv2['path', 'img']() \
      .towhee.clip['data', 'vec'](name='ViT-B/32', modality='image') \
      .select['data', 'vec']() \
      .show()

towhee.dc(["a dog"]) \
      .select['img', 'vec']() \
      .towhee.clip['data', 'vec'](name='ViT-B/32', modality='image') \
      .select['data', 'vec']() \
      .show()
```
<img src="https://towhee.io/towhee/clip/raw/branch/main/tabular1.png" alt="result1" style="height:60px;"/>


<br />



## Factory Constructor

Create the operator via the following factory method

***clip(name, modality)***

**Parameters:**

​   ***name:*** *str*

​   The model name of CLIP. 

​   ***modality:*** *str*

​   Which modality(*image* or *text*) is used to generate the embedding. 

<br />



## Interface

An image-text embedding operator takes a [towhee image](link/to/towhee/image/api/doc) or string as input and generate an image embedding in ndarray.


**Parameters:**

​	***data:*** *towhee.types.Image (a sub-class of numpy.ndarray)*  or *str*

​  The data(image or text based on choosed modality) to generate the embedding.	



**Returns:** *numpy.ndarray*

​   The data embedding extracted by model.