Method and apparatus for extracting facial feature, and method and apparatus for facial recognition
AssigneeSamsung Electronics Co., Ltd.
InventorsNamjoon Kim, Byungin Yoo, Changkyo Lee, Jaejoon Han
Life of this patent
Estimating the term…
Abstract
A method and an apparatus for extracting a facial feature and a method and an apparatus for recognizing a face are provided, in which the apparatus for extracting a facial feature may extract facial landmarks from a current input image, sample a skin region and a facial component region based on the extracted facial landmarks, generate a probabilistic model associated with the sampled skin region and the facial component region, extract the facial component region from a face region included in the input image using the generated probabilistic model, and extract facial feature information from the extracted facial component region.
Claims19
A method of extracting a facial feature, the method comprising:
- extracting facial landmarks from an input image, the input image including a face region;
- generating a probabilistic model associated with a skin region based on the extracted facial landmarks, the generating the probabilistic model including,
- determining a skin color probability distribution of the input image,
- determining at least two facial component color probability distributions of the input image, and
- generating the probabilistic model based on the skin color probability distribution and the at least two facial component color probability distributions;
- extracting a facial component region from the face region using the probabilistic model, the extracting including,
- generating at least one of a skin-excluded probability map for the facial component region and a facial component mask indicating the facial component region based on a result of a comparison between a probability that a current pixel of the input image is a part of the facial component region and a probability that the current pixel is a part of the skin region;
- and
- extracting facial feature information from the extracted facial component region.
A method of recognizing a face, the method comprising:
- extracting facial landmarks from an input image, the input image including a face region;
- generating a probabilistic model associated with a skin region and a facial component region based on the extracted facial landmarks, the generating the probabilistic model including,
- determining a skin color probability distribution of the input image,
- determining at least two facial component color probability distributions of the input image, and
- generating the probabilistic model based on the skin color probability distribution and the at least two facial component color probability distributions;
- extracting facial feature information regarding the facial component region based on the probabilistic model, the extracting including,
- generating at least one of a skin-excluded probability map for the facial component region and a facial component mask indicating the facial component region based on a result of a comparison between a probability that a current pixel of the input image is a part of the facial component region and a probability that the current pixel is a part of in the skin region;
- and
- recognizing a face in the input image based on the extracted facial feature information.
An apparatus for extracting a facial feature, the apparatus comprising:
- a memory storing computer-readable instructions;
- and
- at least one processor configured to execute the computer-readable instructions to,
- extract facial landmarks from an input image, the input image including a face region,
- determine a skin color probability distribution of the input image,
- determine at least two facial component color probability distributions of the input image,
- generate a probabilistic model associated with a skin region based on the extracted facial landmarks, the skin color probability distribution and the at least two facial component color probability distribution,
- extract a facial component region from the face region using the probabilistic model,
- generate at least one of a skin-excluded probability map for the facial component region and a facial component mask indicating the facial component region based on a result of a comparison between a probability that a current pixel of the input image is a part of the facial component region and a probability that the current pixel is a part of the skin region, and
- extract facial feature information from the extracted facial component region.
Dependencies are inferred from the claim text.