Computational Models of Kinship Verification
People
Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen
Abstract
Face recognition is an active research area in the computer vision community. Classical algorithms like Eigenface and Fisherface are known to provide a robust framework for face recognition, while some recent works address the illumination and orientation invariant recognition. However the recognition of people bonded by kinship using facial images is a marginally explored research field. This has led to computational models of kinship verification.
In this work, we present a computational model for kinship verification using novel feature extraction and selection methods, automatically classifying pairs of face images as "related" or "unrelated" (in terms of kinship). First, we conducted a controlled online search to collect frontal face images of 150 pairs of public figures and celebrities, along with images of their parents or children. Next, we propose and evaluate a set of low-level image features for this classification problem. After selecting the most discriminative inherited facial features, we demonstrate a classification accuracy of 70.67% on a test set of image pairs using K-Nearest-Neighbors. Finally, we present an evaluation of human performance on this problem.
Publications
Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen. Towards Computational Models of Kinship Verification. IEEE International Conference on Image Processing, Hong Kong, September 2010 (ICIP '10)
[pdf] [slides] [BibTex]
Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen. Kin Recognition and Ancestry Tracing on Facial Images. Cornell Engineering Research Conference, Ithaca, NY. April 2010 (CERC '10). [poster]
Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen. Face Tracer: A Search Engine for Family Photo Collection with Faces. Fourth North East Student Colloquium on Artificial Intelligence, Amherst, MA. March 2010 (NESCAI '10)
Demos
Sherman Ng, Chang-Hung Lai, Meng-Chen Tsai Advisors: Ruogu Fang, Tsuhan Chen. Automatic Kinship Recognition Using Social Contextual Information. Bits On Our Minds, Ithaca, NY. March 2011. (BOOM'11). [poster]
Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen. Kin Recognition for Family Photo Album Organization and Live Demonstration. Demo session at European Conference on Computer Vision, Crete, Greece, September 2010 (ECCV'10).
Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen. Automatic Kinship Verification. Bits On Our Minds, Ithaca, NY. February 2010. (BOOM'10). [poster]
Dataset
We introduce the first comprehensive publicly available kinship database of 143 pairs of parents and children (300 cropped frontal face images of 100 by 100 pixels). Kinship Verfication. (Note that 7 families are removed from the original dataset of 150 families due to privacy issue.)
The images below gives a glimpse of the parents and children pairs in the dataset. Parents dataset on the left and children dataset on the right.
If you use the 'Kinship verfication dataset' in your research, please cite the following paper,
- Ruogu Fang, Kevin D. Tang, Noah Snavely, Tsuhan Chen. "Towards Computational Models of Kinship Verification", IEEE International Conference on Image Processing, 2010.
Please inform us about your accuracies on this dataset, so that we may make them publicly available for easy comparison and citation.
