vulkan VS pigo

Unfortunately Gorgonia (and thus Goro) only supports CPU and CUDA, but there is a Vulkan branch in development. Seems it won't support OpenCL but if Vulkan works then that's probably OK.

Then I saw this library https://github.com/esimov/pigo but as far as I know, this library only detects if an image has face in it, but cannot really classify the face. CMIIW.

The task consists of porting a small portion of the PIGO face detection GO library (https://github.com/esimov/pigo/commit/29b25278c3e3436416440404b95cd3c18e145b3b) especially the face landmarks extraction section to WebAssembly/JavaScript or plain C depending of the programming language your are proficient with.

I tried https://github.com/ageitgey/face_recognition and https://github.com/esimov/pigo so far and they both work OK, but from there, you would need to add EXIF tags into your pictures, so they can be picked up by img-DB.

I'll use this post / opportunity to ask the HN community

(I know this is a technically complicated, and potentially sensitive subject)

I've been approached by a few people who are trying to make prescription glasses less expensive. It's a mix of (licensed) opticians, people interested in offering community health services, etc.

What is the current state of the art for PD measurement based on face / pupil tracking / detection using laptop webcams or mobile (front) cameras?

I imagine (obviously?) that there must be scholar research on the acceptable error margins for a PD measurement (depending on the type of vision condition, i.e. farsightedness, etc.)?

Would using something like Handsfree or https://github.com/esimov/pigo (which has pupil detection) be a good start, or would these be an ~order of magnitude off in terms of the necessary margins?

Thanks a lot.