作者: Alexander Bronstein , Michael Bronstein , Ron Kimmel
DOI:
关键词:
摘要: A method, for electronic face recognition using geodesic representations, of processing such a representation having mouth and being able to adopt open-mouthed or closed states, so that the works regardless whether in is open closed. The method comprises: identifying region within face; applying topological constraint region, representing with constraint. chosen render invariant state