A Powerful Face Preprocessing For Robust Kinship Verification based Tensor Analyses

This academic article discusses a new system developed for kinship verification using facial photographs, an area of research within the field of computer vision. The system preprocesses face images using a Retinex filter and an ellipse mask, extracts features with hist-Gabor wavelets, and applies an efficient dimensionality reduction method called TXQDA. The system then determines whether two people are related. The system was tested using the Cornell KinFace benchmark database, and results demonstrated that it is robust and efficient compared to other strategies currently in use.

Publication date: 19 Dec 2023
Project Page: not specified
Paper: https://arxiv.org/pdf/2312.11290