A method may include obtaining a base facial image, and obtaining a first set of base facial features within the base facial image, the first set of base facial features associated with a first facial AU to be detected in an analysis facial image. The method may also include obtaining a second set of base facial features within the base facial image, the second set of facial features associated with a second facial AU to be detected. The method may include obtaining the analysis facial image, and applying a first image normalization to the analysis facial image using the first set of base facial features to facilitate prediction of a probability of the first facial AU. The method may include applying a second image normalization to the analysis facial image using the second set of base facial features to facilitate prediction of a probability of the second facial AU.

Systems and methods are disclosed herein for determining a diagnosis based on a base skin tone of a patient. In an embodiment, the system receives a base skin tone image of a patient, generates a calibrated base skin tone image by calibrating the base skin tone image using a reference calibration profile, and determines a base skin tone of the patient based on the calibrated base skin tone image. The system receives a concern image of a portion of the patient's skin, and selects a set of machine learning diagnostic models from a plurality of sets of candidate machine learning diagnostic models based on the base skin tone of the patient, each of the sets of candidate machine learning diagnostic models trained to receive the concern image and output a diagnosis of a condition of the patient.

Systems and methods are disclosed herein for determining a diagnosis based on a base skin tone of a patient. In an embodiment, the system receives a base skin tone image of a patient, generates a calibrated base skin tone image by calibrating the base skin tone image using a reference calibration profile, and determines a base skin tone of the patient based on the calibrated base skin tone image. The system receives a concern image of a portion of the patient's skin, and selects a set of machine learning diagnostic models from a plurality of sets of candidate machine learning diagnostic models based on the base skin tone of the patient, each of the sets of candidate machine learning diagnostic models trained to receive the concern image and output a diagnosis of a condition of the patient.

An eye open or closed state detection method, in addition to an eye opening feature, an iris shape feature and/or a vertical direction feature are added to identify the eye open or closed state, where the eye opening feature is used to represent an eye opening degree, the iris shape feature is used to represent a shape of an iris of an eye, and the vertical direction feature is used to represent a change degree of an eyelid curve.

An eye open or closed state detection method, in addition to an eye opening feature, an iris shape feature and/or a vertical direction feature are added to identify the eye open or closed state, where the eye opening feature is used to represent an eye opening degree, the iris shape feature is used to represent a shape of an iris of an eye, and the vertical direction feature is used to represent a change degree of an eyelid curve.

An image processing apparatus includes an analysis unit for analyzing an image, an extraction unit for extracting a reference image corresponding to a specific object, and a first correlation calculation unit and a second correlation calculation unit for presuming an area having a high correlation in an input image by using the reference image. The first correlation calculation unit executes presumption based on components including a direct current (DC) component, and the second correlation calculation unit executes presumption based on an alternate current (AC) component, from which the DC component is eliminated. Based on at least any one of a detection result of the specific object, a feature amount of the specific object, and an analysis result of an imaging scene, acquired by the analysis unit, the first correlation calculation unit and the second correlation calculation unit are switched when the area is presumed.

An image processing apparatus includes an analysis unit for analyzing an image, an extraction unit for extracting a reference image corresponding to a specific object, and a first correlation calculation unit and a second correlation calculation unit for presuming an area having a high correlation in an input image by using the reference image. The first correlation calculation unit executes presumption based on components including a direct current (DC) component, and the second correlation calculation unit executes presumption based on an alternate current (AC) component, from which the DC component is eliminated. Based on at least any one of a detection result of the specific object, a feature amount of the specific object, and an analysis result of an imaging scene, acquired by the analysis unit, the first correlation calculation unit and the second correlation calculation unit are switched when the area is presumed.

Methods and systems for generating image capture device parameter suggestions that would produce an image in, or closer to, a desired aesthetic style. In particular, the systems and methods described herein include pattern recognition techniques which are utilized to extract visual features from a set of images, those features defining an aesthetic style. The set of images comprise exemplars of said aesthetic style as well as images representing a plurality of variations in image capture parameters. An algorithm is trained to discriminate between exemplar and variation images based on their extracted visual features. When the system is presented with an input image the same visual features are extracted from it and are compared to the characteristic visual features of the exemplar and variation images with the trained discriminator algorithm. The similarity of the input image to exemplar and variation images are used to generate image capture device parameter suggestions.

A machine learning method operates on training instances from a plurality of domains including one or more source domains and a target domain. Each training instance is represented by values for a set of features. Domain adaptation is performed using stacked marginalized denoising autoencoding (mSDA) operating on the training instances to generate a stack of domain adaptation transform layers. Each iteration of the domain adaptation includes corrupting the training instances in accord with feature corruption probabilities that are non-uniform over at least one of the set of features and the domains. A classifier is learned on the training instances transformed using the stack of domain adaptation transform layers. Thereafter, a label prediction is generated for an input instance from the target domain represented by values for the set of features by applying the classifier to the input instance transformed using the stack of domain adaptation transform domains.

A method for evaluating an individual fingerprint template by using a remote dataset, comprising the steps of capturing a fingerprint representation by a fingerprint reader on a device, extracting significant data from the captured fingerprint representation, thereby creating an individual fingerprint template for the captured fingerprint representation, transmitting the individual fingerprint template from the device to a remote dataset comprising a plurality of fingerprint templates, determining an impostor score distribution for the individual fingerprint template, determining a security threshold for the individual fingerprint template, and transmitting the determined security threshold to the device. The advantage of the invention is that an individual security threshold can be set for a user, which will improve the FAR distribution of a device.