TrueFace.ai hackathon in 500 startups SF
I believe it was my second time coming to 500 SF office and first ever hackathon. It was held by TrueFace.ai which is a part of batch 22. The fact that face recognition is easily available just by making an API call is pretty exiting, I must say. As I understood, one of the key features of the API is a spoof detection. Image that contains an image of a person’s face will not pass the test.

You can detect if there is a face or multiple faces on a picture calling “detect”…

For face recognition need to create a collection first…

Then enroll at least 5 pictures with the same person for training…

When images are there you need to update a collection so system will know there is a new data loaded…

Next step is to train the collection classifier…

And, finally, after system is trained, you can identify a particular user by sending his/her image…

Or you can match a person with all existing profiles and get the score for each profile.

