Search papers, labs, and topics across Lattice.
This paper addresses the problem of face recognition with only a single labeled image of the target person and a stream of unlabeled data. They formulate the problem as one-class classification and propose a non-parametric algorithm to learn a face model from the single labeled image and the unlabeled stream. Experiments on a dataset of 43 people demonstrate a 25+% improvement in recall compared to baselines, achieving 90% recognition rate with near-zero false positives.
Unlock 25% better face recognition from a single labeled image by leveraging readily available unlabeled data streams.
Face recognition from a single image per person is a challenging problem because the training sample is extremely small. We consider a variation of this problem. In our problem, we recognize only one person, and there are no labeled data for any other person. This setting naturally arises in authentication on personal computers and mobile devices, and poses additional challenges because it lacks negative examples. We formalize our problem as one-class classification, and propose and analyze an algorithm that learns a non-parametric model of the face from a single labeled image and a stream of unlabeled data. In many domains, for instance when a person interacts with a computer with a camera, unlabeled data are abundant and easy to utilize. This is the first paper that investigates how these data can help in learning better models in the single-image-per-person setting. Our method is evaluated on a dataset of 43 people and we show that these people can be recognized 90% of time at nearly zero false positives. This recall is 25+% higher than the recall of our best performing baseline. Finally, we conduct a comprehensive sensitivity analysis of our algorithm and provide a guideline for setting its parameters in practice.
