1

I am able to do supervised learning for Facial Recognition with TensorFlow and especially after reading this great tutorial. I highly recommend the tutorial even if you don't intend on running the code.

Does anyone know of an open sourced library that does clustering of images based on facial recognition similar to Google Photos for instance? Or even a tutorial with sample code?

I found a good tutorial on autoencoders, so building my own is an option, but I wanted to check if there's something open sourced, or some tutorial that has code I maybe able to borrow?

Sam Hammamy
  • 10,819
  • 10
  • 56
  • 94
  • openface can do that. Why? Because it's CNN is learning a metric-embedding (euclidean) of faces. (did only skim through your clustering link, but that does not look like clustering!) – sascha Sep 16 '17 at 14:52
  • Similarity detection is probably more accurate than clustering. But the end goal is to group the images by the person in the image – Sam Hammamy Sep 16 '17 at 15:12
  • That's also covered by openface, as similarity is then the (inverse) distance in the metric-space. – sascha Sep 16 '17 at 15:14

0 Answers0