# Which face is which? Persistence faces tracking

**URL:** <https://discourse.processing.org/t/which-face-is-which-persistence-faces-tracking/38282>\
**Category:** Libraries\
**Created:** [August 9, 2022, 2:29pm UTC](https://discourse.processing.org/t/which-face-is-which-persistence-faces-tracking/38282 "2022-08-09T14:29:18Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![AndreasRef](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.processing.org/andreasref/32/133_2.png) [@AndreasRef](https://discourse.processing.org/u/AndreasRef)\
**Post date:** [August 9, 2022, 2:29pm UTC](https://discourse.processing.org/t/which-face-is-which-persistence-faces-tracking/38282/1 "2022-08-09T14:29:18Z")

</div>

Hi!  
I have this p5js sketch that uses tensorflow Blazeface to track faces.

> **[p5.js Web Editor](https://editor.p5js.org/AndreasRef/sketches/ueRpdqRQb)**
>
> A web editor for p5.js, a JavaScript library with the goal of making coding accessible to artists, designers, educators, and beginners.

I was wondering how I would go about keeping track of multiple faces for an interactive installation I am doing?  
Kinda like [Face-It/FaceDetectMemory.pde at master · shiffman/Face-It · GitHub](https://github.com/shiffman/Face-It/blob/master/OpenCV/FaceDetectMemory/FaceDetectMemory.pde) but perhaps even simpler/more roboust and in JavaScript?

I am not looking for actual face recognition btw.