deletions | additions       diff --git a/When_it_comes_to_Automatic__.tex b/When_it_comes_to_Automatic__.tex  index 7d98457..01ae6ec 100644  --- a/When_it_comes_to_Automatic__.tex  +++ b/When_it_comes_to_Automatic__.tex 
...
When it comes to Automatic face analysis which includes ,e.g., face detection, face recognition, gender classification, age estimation, facial expression recognition and kinship verification has become one of the most active topics in computer vision research \citet{jain2005handbook}. Fortunately, the recently introduced low-cost depth sensors such as Microsoft Kinect allow extracting directly 3D information, together with RGB color images. Detecting human face, estimating its pose and tracking it are crucial steps for many applications in computer vision and human–machine interaction. Many research works have shown the usefulness of Kinect for face detection and segmentation [18–20], \citet{masselli2012real},  head pose estimation and normalization \citet{niese2013accurate} and face tracking \citet{li2013head}. Some works \citet{fanelli2011real} make use of depth maps only while others \citet{yang2012face} combine both RGB and depth data. On the other hand, combination of color and depth information for face detection and tracking and head pose estimation is proven to achieve more robustness than using the two modalities separately. \citet{boutellaa2015use}