A dense flow-based framework for real-time object registration under compound motion
作者:
Highlights:
• A moving object often has elastic and deformable surfaces (e.g., a human head).
• Measuring surface deformation while the object itself is also moving is challenging.
• We present a generic video alignment framework to characterize surface deformations.
• The proposed algorithm is suitable for spontaneous expressions recognition.
• It also improves the performance of face image super-resolution.
摘要
Highlights•A moving object often has elastic and deformable surfaces (e.g., a human head).•Measuring surface deformation while the object itself is also moving is challenging.•We present a generic video alignment framework to characterize surface deformations.•The proposed algorithm is suitable for spontaneous expressions recognition.•It also improves the performance of face image super-resolution.
论文关键词:Object registration,Spontaneous facial expression,SIFT flow,Optical flow,Super-resolution
论文评审过程:Received 14 March 2016, Revised 11 October 2016, Accepted 12 October 2016, Available online 14 October 2016, Version of Record 21 October 2016.
论文官网地址:https://doi.org/10.1016/j.patcog.2016.10.015