Final published version
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
}
TY - GEN
T1 - Face Symmetry Analysis Using a Unified Multi-task CNN for Medical Applications
AU - Storey, Gary
AU - Jiang, Richard
PY - 2019/1/1
Y1 - 2019/1/1
N2 - Facial symmetry analysis can provide an important role in the diagnosis and rehabilitation of medical conditions like facial paralysis issues such as bell’s palsy. Recent advances in computer vision techniques specifically the use of deep convolutional neural networks and multi-task learning provide a gateway to fast and state-of-the-art accurate methods for object detection tasks. In this paper, we present a novel unified multi-task CNN framework for simultaneous object proposal, face detection and face symmetry analysis. We highlight the potential possibilities for the use of such a framework within the medical domain through the experimental results on two test data sets. The results are promising showing high level of accuracy for both the task of face detection and symmetry analysis while also highlighting the efficient computational overhead for our proposed method which can process an image in 0.04 s.
AB - Facial symmetry analysis can provide an important role in the diagnosis and rehabilitation of medical conditions like facial paralysis issues such as bell’s palsy. Recent advances in computer vision techniques specifically the use of deep convolutional neural networks and multi-task learning provide a gateway to fast and state-of-the-art accurate methods for object detection tasks. In this paper, we present a novel unified multi-task CNN framework for simultaneous object proposal, face detection and face symmetry analysis. We highlight the potential possibilities for the use of such a framework within the medical domain through the experimental results on two test data sets. The results are promising showing high level of accuracy for both the task of face detection and symmetry analysis while also highlighting the efficient computational overhead for our proposed method which can process an image in 0.04 s.
KW - Computer vision
KW - Face recognition
KW - Face analysis
KW - Medical diagnosis
U2 - 10.1007/978-3-030-01057-7_36
DO - 10.1007/978-3-030-01057-7_36
M3 - Conference contribution/Paper
SN - 9783030010560
T3 - Intelligent Systems and Applications
SP - 451
EP - 463
BT - IntelliSys 2018: Intelligent Systems and Applications
A2 - Arai, Kohei
A2 - Kapoor, Supriya
A2 - Bhatia, Rahul
PB - Springer
CY - Cham
ER -