題目：Cascaded face alignment via intimacy definition \r
feature
報告人：Prof. Kenneth K.M. Lam
Department of Electronic and \r
Information Engineering
The Hong Kong Polytechnic University Hong \r
Kong
邀請人：邱國平教授
時間：2017年11月27日（周一）15：00
地點：深圳大學南區(qū)基礎(chǔ)實驗樓北座信息工程學院N710會議室
摘要：
Recent \r
years have witnessed the emerging popularity of regression-based face aligners, \r
which directly learn mappings between facial appearance and shape-increment \r
manifolds. In this talk, I will introduce a random-forest-based, cascaded \r
regression model for face alignment by using a novel locally lightweight \r
feature, namely intimacy definition feature (IDF). This feature is more \r
discriminative than the pose-indexed feature, more efficient than the histogram \r
of oriented gradients (HOG) feature and the scale-invariant feature transform \r
(SIFT) feature, and more compact than the local binary feature (LBF). \r
Experimental validation of the algorithm shows that it achieves state-of-the-art \r
performance when testing on some challenging datasets. Compared with the \r
LBF-based algorithm, the method achieves about twice the speed, 20% improvement \r
in terms of alignment accuracy, and save an order of magnitude on memory \r
requirement.
歡迎各位老師和同學參加。
