Digital Image Animation

• US 10,198,846 B2
• Filed: 08/22/2016
• Issued: 02/05/2019
• Est. Priority Date: 08/22/2016
• Status: Active Grant

• Current Assignee: Adobe Inc.
• Original Assignee: Adobe Inc.
• Inventors: Carr, Nathan Aaron
• Primary Examiner(s): Wills, Diane M
• Application Number:
  US15/243,130
• Publication Number:
  US 20180053333A1
• Time in Patent Office: 897 Days
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06T 13/80 2D [Two Dimensional] animation, e.g. using sprites

Abstract

Digital image animation techniques are described. In one example, animations are used for a single digital image to permit movement or other effects to be exhibited as part of the digital image without requiring multiple frames as in conventional techniques. Transformation of the single digital image by the animations may also be coordinated, such as to synchronize or not synchronize movement of objects to promote realism. In another example, portions and even an entirety of these techniques may be performed automatically and without user intervention. Machine learning, for instance, may be employed using a neural network to classify the digital image into one or more semantic classes. The semantic classes may then be used to recommend animations to transform the digital image.

Claims

1. In a digital medium environment, a system comprising:

an animation selection module implemented at least partially in at least one computing device including one or more processors and one or more computer readable storage media storing instructions that, when executed by the one or more processors, perform operations including;

obtaining, by the at least one computing device, a digital image;

classifying, by the at least one computing device, the digital image into at least one semantic class using a neural network as part of machine learning, the neural network trained to recognize the at least one semantic class;

generating, by the at least one computing device, a recommendation of a first animation and a second animation from a plurality of animations based on the at least one semantic class;

selecting, by the at least one computing device, the first animation and the second animation from the plurality of animations to be applied to the digital image based on classifying the digital image into the at least one semantic class; and

a digital content transformation module implemented at least partially in hardware of the computing device to perform operations including transforming an appearance of the digital image based on the selection, the transforming including coordinating;

a timing of output of the first animation and the second animation to each other within an iteration; and

a timing of the output of the first and second animations to each other to give a different appearance between multiple said iterations of the first and second animations.

2. The system as described in claim 1, wherein the classifying and the generating are performed automatically and without user intervention using a neural network.

3. The system as described in claim 1, wherein the digital content transformation module is further configured to transform the appearance of the digital image, automatically and without user intervention, based on the selection of the first animation and the second animation.

4. In a digital medium environment, a method implemented by at least one computing device, the method comprising:

obtaining, by the at least one computing device, a digital image;

classifying, by the at least one computing device, the digital image into at least one semantic class, the classifying performed using a neural network as part of machine learning, the neural network trained to recognize the at least one semantic class;

selecting, by the at least one computing device, first and second animations from a plurality of animations to be applied to the digital image based on classifying the digital image into the at least one semantic class; and

transforming, by the at least one computing device, an appearance of the digital image based on the selected first and second animations, the transforming including coordinating;

a timing of output of the first animation and the second animation to each other within an iteration; and

a timing of the output of the first and second animations to each other to give a different appearance between multiple said iterations of the first and second animations.

5. The method as described in claim 4, wherein the selection is further based on a user input received via a user interface output by the at least one computing device.

6. The method as described in claim 4, wherein the selecting based at least in part on machine learning includes:

classifying, by the at least one computing device, the digital image into the at least one semantic class using machine learning; and

generating, by the at least one computing device, a recommendation of the first and second animation from the plurality of animations based on the at least one semantic class.

7. The method as described in claim 6, wherein the machine learning is performed by the at least one computing device using a neural network trained for the at least one semantic class.

8. The method as described in claim 4, wherein the at least one semantic class includes identification of an object included in the digital image or a type of scene of the digital image.

9. The method as described in claim 4, wherein coordinating the timing of the first animation and the second animation within the iteration includes synchronizing actions performed by an object as part of the first animation with actions performed by another object as part of the second animation.

10. The method as described in claim 4, wherein coordinating the timing of output of the first animation and the second animation within the iteration includes configuring actions performed by an object as part of the first animation as not synchronized with actions performed by another object as part of the second animation.

11. The method as described in claim 4, wherein the transforming includes:

segmenting, by the at least one computing device, the digital image into a plurality of segments; and

replacing at least one of the plurality of segments having an object with at least one of the selected first and second animations.

12. In a digital medium environment, a system comprising:

means for generating a recommendation of first and second animations from a plurality of animations based on at least one semantic class identified from a digital image using machine learning;

means for selecting the first animation and the second animation from the plurality of animations to be applied to the digital image based on the identified at least one semantic class; and

means for transforming the digital image for output in an immersive augmented or virtual reality environment using the first animation and the second animation and output the transformed digital image as including the first animation and the second animation, the transforming including coordinating;

a timing of output of the first animation and the second animation to each other within an animation; and

a timing of the output of the first and second animations to each other to give a different appearance between multiple said iterations of the first animation and the second animation.

13. The system as described in claim 12, wherein the machine learning is performed using a neural network.

14. The system as described in claim 12, wherein the transforming means is configured to transform the digital image by:

segmenting the digital image into a plurality of segments; and

replacing at least one of the plurality of segments with at least one of the first animation and the second animation.