Tracking and recognition of faces using selected region classification

摘要

Methods, apparatuses, and articles associated with facial tracking and recognition are disclosed. In embodiments, facial images may be detected in video or still images and tracked. After normalization of the facial images, feature data may be extracted from selected regions of the faces to compare to associated feature data in known faces. The selected regions may be determined using a boosting machine learning processes over a set of known images. After extraction, individual two-class comparisons may be performed between corresponding feature data from regions on the tested facial images and from the known facial image. The individual two-class classifications may then be combined to determine a similarity score for the tested face and the known face. If the similarity score exceeds a threshold, an identification of the known face may be output or otherwise used. Additionally, tracking with voting may be performed on faces detected in video. After a threshold of votes is reached, a given tracked face may be associated with a known face.

主权项

1. A method comprising:
tracking, by a computer apparatus, a facial image with varying appearances in a plurality of image frames of a video signal; and identifying, by the computer apparatus, a known facial image as potentially associated with the tracked facial image with varying appearances upon meeting an identification condition, wherein the identifying comprises:
determining, by the computer apparatus, a set of candidate regions from a plurality of areas, wherein the plurality of areas spans an entirety of the tracked facial image with varying appearances, and wherein each candidate region of the set of candidate regions comprises varying combinations of the areas,selecting, by the computer apparatus, a set of facial regions from the set of candidate regions based on a result of a machine learning process,extracting, by the computer apparatus, facial feature data from a subset of facial regions of the tracked facial image with varying appearances, anddetermining, by the computer apparatus, whether a number of associations of the tracked facial image with varying appearances to a number of known facial images, based on the extracted facial feature data in the subset of facial regions, exceeds a predetermined threshold across the varying appearances, and wherein the identification condition is based at least in part on the determined associations.