Why you use three steps's threshold [0.6, 0.7, 0.7] for detection? Thank you!!!

bearsprogrammer / real-time-deep-face-recognition

using facenet algorithm

MIT License

233 stars 100 forks source link

Open mactiendinh opened 6 years ago

mactiendinh commented 6 years ago

And in Facenet paper, they use threshorld = 1.1 to determine the detected threshold. Does Your threshold and their threshold have different?

lilmangolil commented 4 years ago

This threshold is usd for MTCNN,not facenet.