BSIF Features Learning using TXQEDA Tensor Subspace for kinship verification

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Echahid Cheikh Larbi Tebessi university
Facial kinship verification is a hard research domain in vision that has very interesting regard in the latest decennial. Various applications were really realized in social media, biometrics, and development in studies of demographic. But the result accuracies obtained that is so weak to predict kinship relationships by facial appearance. To take up this challenge and tackle this problem. We use a new approach called Color BSIF learning an approach that has appeared as an encouraging solution. The aim is to solve problem KV by using the color BSIF learning features with the TXQEDA method for dimensionality reduction and data classification in order to train the model, Let's test the kinship facial verification application namely the Cornell Kinface database. This framework ameliorates the time cost and efficiency. The experimental results obtained surpass other states of the art methods
Verification Kinship, Color BSIF learning, TXQEDA, Cornell Kinship database, Facial kinship verification