Techniques are provided for multi-modal sensitive recognition. A digital data set for an object is obtained according to a modality, where the digital data set includes digital representations of the object at different values of a dimension of relevance of the modality. A reference location associated with the object is identified. A modal descriptor is derived for the modality according to an implementation of a multi-modal recognition algorithm by deriving a set of feature descriptors for the reference location and at the different values of the corresponding dimension of relevance, calculating a set of differences between the feature descriptors in the set of feature descriptors, and aggregating the set of differences into the modal descriptor. A device is then configured to initiate an action as a function of the modal descriptor.

In computation of a line-of-sight direction in a control unit, a first computed line-of-sight direction is obtained from the center of a pupil and a reflection point of cornea reflection light, a second computed line-of-sight direction is obtained by computing the ellipse shape of an iris or the pupil, and a third computed line-of-sight direction is obtained on the basis of a face orientation and the relative position of the iris or pupil in an eye. The line-of-sight direction can be computed over a wide range with high accuracy by selecting one of the computed line-of-sight directions.

Contextual parameters of digital media content may be obtained. The digital media content may be associated with a content capture user and/or an end user. Editing parameters having values defining one or more editing attributes, including one or more selected moments of interest, of an edited version of the digital media content may be received. Individual post-capture user profiles may include expertise attributes associated with individual post-capture users. A set of post-capture users may be identified as potential matches for creating the edited version of the digital media content based upon one or more of values of contextual parameters, editing parameters, one or more expertise attributes of the post-capture user profiles, and/or other information. The set of post-capture users may be presented to the content capture user and/or the end user for selection of one of the post-capture users to create the edited version of the digital media content.

The present invention discloses a multispectral stereo camera self-calibration algorithm based on track feature registration, and belongs to the field of image processing and computer vision. Optimal matching points are obtained by extracting and matching motion tracks of objects, and external parameters are corrected accordingly. Compared with an ordinary method, the present invention uses the tracks of moving objects as the features required for self-calibration. The advantage of using the tracks is good cross-modal robustness. In addition, direct matching of the tracks also saves the steps of extraction and matching the feature points, thereby achieving the advantages of simple operation and accurate results.

Exemplary embodiments are directed to a system of enhancing biometric analysis matching. A processing device is configured to analyze the probe short-range and broadband iris texture information of a probe image for iris biometric authenticity, and based on the biometric authenticity of the probe short-range iris texture information and the probe broadband iris texture information, determine the biometric authenticity of the subject. Exemplary embodiments are also directed to a system of enhancing biometric analysis matching efficiency. The processing device generates an optimized order of enrollment iris biometric data includes a listing of the enrollment iris biometric data ordered by closest match to furthest match between the probe and enrollment broadband iris texture information. The processing device analyzes the iris biometric data for biometric authenticity starting with the closest match between the probe and enrollment broadband iris texture information.

System and related methods for applying machine learning to the classification of surface materials using images of spots of lights, such as resulting from a laser beam impinging the surface. A classifier trained using such spot images, resulting from light beams imping the surface, achieves excellent classification results, in spite of a lack of fine surface details in these images as compared to a more uniformly lit larger scene that would appear to contain more information on the surface type. Classifiers can achieve classification accuracies on biological tissues significantly above 90% using a number of well-known classifier architectures. The classification results can be used to generate a map of classified surface types and the combination of such with a three-dimensional model of a surface having classified surface portions reconstructed from a pattern of spots projected onto the surface.

Apparatuses and methods of operating apparatuses are disclosed. An apparatus comprises a flexible substrate and circuitry fabricated on the flexible substrate to perform data processing. At least one strain detector generates a strain signal which is dependent on a flexing state of the strain detector on the flexible substrate. A strain history control unit samples the at least one strain signal from the at least one strain detector at a plurality of time points and records a strain snapshot at each time point comprising data dependent on the at least one strain signal from the at least one strain detector. The data processing performed by the circuitry is dependent on the plurality of strain snapshots recorded

There is provided an imaging system. The imaging system comprising a multispectral camera configured to capture a multispectral image of an object, an RGB camera configured to capture a color image of the object, at least one storage device configured to store spectrum information for each of a plurality of labeled objects, and processing circuitry. The processing circuitry is configured to determine, based on the captured multispectral image, spectrum information associated with the object, associate, based at least in part, on the spectrum information associated with the object and the stored spectrum information for each of the plurality of objects, a label with the color image of the object, and store, on the at least one storage device, the color image and the associated label as training data.

This disclosure relates to detecting hand gesture input using an electronic device, such as a wearable device strapped to a wrist. The device can have multiple photodiodes, each sensing light at a different position on a surface of the device that faces skin of a user. Examples of the disclosure detect hand gestures by recognizing patterns in sensor data that are characteristic of each hand gesture, as the tissue expands and contracts and anatomical features in the tissue move during the gesture.

A facial recognition method includes receiving, by an electronic device, a first instruction triggering the electronic device to perform facial recognition, emitting in response to the first instruction, infrared light with a light spot using an infrared projector, collecting first image information of a first object using a first camera, collecting second image information of the first object using a second camera, calculating depth information of the first object based on the first image information, the second image information, the first length, a lens focal length of the first camera, and a lens focal length of the second camera, and performing user identity verification on the first object using the first image information and the depth information of the first object.