As shown in Section 3, any cascade shape indexed face alignment method [4,
28, 29, 21] can actually be used for cascade detection. We adopt the most recent
and state-of-the-art approach [21], because it is most accurate, fastest, and easy
to integrate the weak classifier learning for detection. Notably, it is significantly
faster (dozens of times) than the previous best methods [4, 28] and runs in thousands
of frames per second to align dozens of facial points. Such high speed is
crucial for a real time face detector as well
As shown in Section 3, any cascade shape indexed face alignment method [4,28, 29, 21] can actually be used for cascade detection. We adopt the most recentand state-of-the-art approach [21], because it is most accurate, fastest, and easyto integrate the weak classifier learning for detection. Notably, it is significantlyfaster (dozens of times) than the previous best methods [4, 28] and runs in thousandsof frames per second to align dozens of facial points. Such high speed iscrucial for a real time face detector as well
การแปล กรุณารอสักครู่..
