| Summary: | 碩士 === 國立清華大學 === 資訊工程學系 === 101 === Different persons usually exhibit various appearance changes when posing the same expression. This person-dependent behavior often complicates automatic facial expression recognition. In this thesis, to address the person-independent expression recognition problem, we propose a Dual Subspace Nonnegative Graph Embedding (DSNGE) to represent expressive images using two subspaces: identity and expression subspaces. The identity subspace characterizes person-dependent appearance variations; whereas the expression subspace characterizes person-independent expression variations. With DSNGE, we decompose each facial image into an identity part and an expression part represented by their corresponding nonnegative bases. We also address the intra-class variations issue in the expression recognition problem, and further devise a graph-embedding constraint on the expression subspace to tackle this problem. Our experimental results show that the proposed DSNGE outperforms other graph-based nonnegative factorization methods and existing expression recognition methods on CK+, JAFFE and TFEID databases.
|
|---|
