mobilefacenet/README.md

# MobileFaceNet Face Landmark Detector

*author: David Wang*


<br />



## Description

[MobileFaceNets](https://arxiv.org/pdf/1804.07573) is a class of extremely efficient CNN models to extract 68 landmarks from a facial image. It use less than 1 million parameters and is specifically tailored for high-accuracy real-time face verification on mobile and embedded devices. This repository is an adaptation from [cuijian/pytorch_face_landmark](https://github.com/cunjian/pytorch_face_landmark).


<br />


## Code Example

Extract facial landmarks from './img1.jpg'. 

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

```python
from towhee import pipe, ops, DataCollection

p = (
    pipe.input('path')
        .map('path', 'img', ops.image_decode())
        .map('img', 'landmark', ops.face_landmark_detection.mobilefacenet())
        .output('img', 'landmark')
)

DataCollection(p('./img1.jpg')).show()
```

<img src="https://towhee.io/face-landmark-detection/mobilefacenet/raw/branch/main/result.png" alt="result1" style="height:20px;"/>


<br />



## Factory Constructor

Create the operator via the following factory method:

***face_landmark_detection.mobilefacenet(pretrained = True)***

**Parameters:**

​  ***pretrained***

​  whether load the pre-trained weights.

​  supported types: `bool`, default is True, using pre-trained weights.


<br />



## Interface

An image embedding operator takes an image as input. it extracts the embedding as ndarray.

**Parameters:**

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

​	The input image.


**Returns:** *numpy.ndarray*	

​	The extracted facial landmarks.





# More Resources

- [Approximate Nearest Neighbors Oh Yeah (Annoy) - Zilliz blog](https://zilliz.com/learn/approximate-nearest-neighbor-oh-yeah-ANNOY): Discover the capabilities of Annoy, an innovative algorithm revolutionizing approximate nearest neighbor searches for enhanced efficiency and precision.
- [How to Get the Right Vector Embeddings - Zilliz blog](https://zilliz.com/blog/how-to-get-the-right-vector-embeddings): A comprehensive introduction to vector embeddings and how to generate them with popular open-source models.
- [Hugging Face Inference Endpoints & Zilliz Cloud](https://zilliz.com/product/integrations/hugging-face): nan
- [Transforming Text: The Rise of Sentence Transformers in NLP - Zilliz blog](https://zilliz.com/learn/transforming-text-the-rise-of-sentence-transformers-in-nlp): Everything you need to know about the Transformers model, exploring its architecture, implementation, and limitations
- [Understanding ImageNet: A Key Resource for Computer Vision and AI Research](https://zilliz.com/glossary/imagenet): The large-scale image database with over 14 million annotated images. Learn how this dataset supports advancements in computer vision.
- [Build a Multimodal Search System with Milvus - Zilliz blog](https://zilliz.com/blog/how-vector-dbs-are-revolutionizing-unstructured-data-search-ai-applications): Implementing a Multimodal Similarity Search System Using Milvus, Radient, ImageBind, and Meta-Chameleon-7b
- [Zilliz-Hugging Face partnership - Explore transformer data model repo](https://zilliz.com/partners/hugging-face): Use Hugging Face’s community-driven repository of data models to convert unstructured data into embeddings to store in Zilliz Cloud, and access a code tutorial.