Cascaded regression with sparsified feature covariance matrix for facial landmark detection
详细信息 查看全文
- 作者：Enrique Sá ; nchez-Lozano ; Brais Martinez ; Michel F. Valstar ; Michel.Valstar@nottingham.ac.uk" class="auth_mail" title="E-mail the corresponding author ;
- 关键词：Supervised descent method ; Cascaded regression ; Facial point localisation
- 刊名：Pattern Recognition Letters
- 出版年：2016
- 出版时间：1 April 2016
- 年：2016
- 卷：73
- 期：Complete
- 页码：19-25
- 全文大小：2059 K
文摘
- •
We analyse the value of context in the Supervised Descent Method (SDM).
- •
We found that the use of the full context is not beneficial in general, only helping in certain situations.
- •
We highlight the relation between context and data variance context.
- •
We show that optimal performance results from discarding some context in a greedy manner.