"Time flies like an arrow; fruit flies like a banana"

5 comments

  • Woah, I was expecting so much of a result. Great tests!

    PS: there’s a small typo in the author’s Twitter link. It only takes your Twitter account name and not the twitter URL 😉

    • Thanks a lot. 🙂 Your question really got me thinking, so I explored the matter somewhat in depth.

      And thanks again for the notice about my Twitter URL, fixed it. 😉

    • In terms of resolution, Face API can detect faces down to 36×36 pixels images, and black and white is not an issue as far as I know.
      But in terms of the quality of the photograph itself, blur, defects, noise and so on can in fact affect the outcome. Did not really take that into consideration here. 🙂

      In a more realistic scenario however, you can also request to the Face API to detect features like blur (how much face is blurred), exposure (good or bad exposure of face) and noise levels in pictures during the Detect request.You could then return an alert in case picture is too blurred or has a bad exposure. You can read more on the Face API reference. 😉

  • Pingback: essayforme

Leave a Reply