Scaled perspective zoom on resource constrained devices
AssigneeSnap Inc.
InventorsLinjie Luo, Chongyang Ma, Zehao Xue
Life of this patent
Estimating the term…
Abstract
A dolly zoom effect can be applied to one or more images captured via a resource-constrained device (e.g., a mobile smartphone) by manipulating the size of a target feature while the background in the one or more images changes due to physical movement of the resource-constrained device. The target feature can be detected using facial recognition or shape detection techniques. The target feature can be resized before the size is manipulated as the background changes (e.g., changes perspective).
Claims20
A method comprising:
- generating, by one or more processors of a user device, an image using an image sensor of the user device;
- generating, using a convolutional neural network, image feature areas for different image features in the image;
- identifying a target feature from one of the image feature areas, the target feature corresponding to a physical object in the image;
- generating, using the image sensor, a zoom video sequence by maintaining the scaling, for each image in the zoom video sequence, of the target feature without maintaining the scaling of image features areas that are not the target feature such that the target feature covers a depiction of the physical object in the zoom video sequence as the depiction of the physical object diminishes in size due to the user device moving away from the physical object;
- and
- storing the zoom video sequence in memory of the user device.
A user device comprising:
- one or more processors;
- an image sensor;
- and
- a memory storing instructions that, when executed by the one or more processors, cause the user device to perform operations comprising
- generating an image using the image sensor;
- generating, using a convolutional neural network, image feature areas for different image features in the image;
- identifying a target feature corresponding to a physical object in the image;
- generating, using the image sensor, a zoom video sequence by maintaining the scaling, for each image in the zoom video sequence, of the target feature without maintaining the scaling of image features areas that are not the target feature such that the target feature covers a depiction of the physical object in the zoom video sequence as the depiction of the physical object diminishes in size due to the user device moving away from the physical object;
- and
- storing the zoom video sequence in the memory.
A non-transitory machine-readable storage device embodying instructions that, when executed by a machine, cause the machine to perform operations comprising:
- generating an image using an image sensor of the machine;
- generating, using a convolutional neural network, image feature areas for different image features in the image;
- identifying a target feature from one of the image feature areas, the target feature corresponding to a physical object in the image;
- generating, using the image sensor, a zoom video sequence by maintaining the scaling, for each image in the zoom video sequence, of the target feature without maintaining the scaling of image features areas that are not the target feature such that the target feature covers a depiction of the physical object in the zoom video sequence as the depiction of the physical object diminishes in size due to the machine moving away from the physical object;
- and
- storing the zoom video sequence.
Dependencies are inferred from the claim text.