Attribute recognition via visual search
AssigneeAdobe Systems Incorporated
InventorsZhe Lin, Jonathan Brandt, Xiaohui Shen
Life of this patent
Estimating the term…
Abstract
One exemplary embodiment involves identifying feature matches between each of a plurality of object images and a test image, each feature matches between a feature of a respective object image and a matching feature of the test image, wherein there is a spatial relationship between each respective object image feature and a test image feature, and wherein the object depicted in the test image comprises a plurality of attributes. Additionally, the embodiment involves estimating, for each attribute in the test image, an attribute value based at least in part on information stored in a metadata associated with each of the object images.
Claims20
A computer-implemented method comprising:
- receiving, at a computing device comprising a processor, a first image depicting a first object and a plurality of second images detected to depict second objects similar to the first object, the first object comprising a plurality of attributes;
- identifying a plurality of attribute values of the second objects depicted in the second images, the attribute values being stored in corresponding metadata associated with the second images;
- and
- for each attribute of the plurality of attributes, generating an estimate for a corresponding attribute value of the first object depicted in the first image based on the attribute values of the second objects stored in the corresponding metadata associated with the second images.
A system comprising:
- a processor for executing instructions stored in computer-readable medium on one or more devices,
- the instructions comprising one or more modules configured to perform the steps comprising:
- receiving a first image having a first face and a plurality of second images, each one of the second images depicting a second face detected to be similar to the first face, each of the first and second faces comprising a plurality of attributes;
- identifying a plurality of second attribute values of attributes of the second faces, the second attribute values being stored in a metadata associated with each respective second image;
- and
- for each attribute of the first face in the first image, generating an estimate for a first attribute value by computing a likely attribute value based on the second attribute values stored in the metadata associated with each of the second images.
A method comprising:
- identifying, by a process of a computing device, feature matches between each of a plurality of second images and a first image, each of the feature matches between a second feature depicted in a respective second image and a first feature depicted in the first image, wherein there is a spatial relationship between each respective second feature and the first feature, and wherein the first feature depicted in the first image comprises a plurality of attributes;
- and
- for each attribute of the first feature depicted in the first image, generating, by the processor, an estimate for an attribute value based at least in part on information stored in metadata associated with matched second features depicted in the second images.
Dependencies are inferred from the claim text.