Granted Patent B2

US 12,456,331 · App. 18/053,641 · Granted Oct 28, 2025

Inventors: Zohreh Azizi (Los Angeles, CA); Surabhi Sinha (San Jose, CA); Siavash Khodadadeh (Mountain View, CA)

Assignee: ADOBE INC.

G06V40/172G06T5/77

Abstract

Claims (59)

1. A method for image generation comprising:

identifying, by a machine learning model, an image generation network that includes an encoder and a decoder;

pruning channels of a first layer of a block of the encoder;

pruning channels of a second layer of the block of the encoder based on the pruned channels of the first layer of the block of the encoder;

pruning channels of a block of the decoder that is connected to the block of the encoder by a skip connection, wherein the channels of the block of the decoder are pruned based on the pruned channels of the block of the encoder;

wherein pruning channels of the block of the decoder comprises:

pruning channels of a first layer of the block of the decoder based on the pruned channels of the first layer of the block of the encoder; and

pruning channels of a second layer of the block of the decoder based on the pruned channels of the second layer of the block of the encoder; and

generating an image using the image generation network based on the pruned channels of the block of the encoder and the pruned channels of the block of the decoder.

2. The method of claim 1 , further comprising:

identifying an input image and a portion of the input image for inpainting; and

inpainting the portion of the input image using the image generation network to obtain an inpainted image.

3. The method of claim 1 , further comprising:

identifying an image of a face; and

generating an anonymized image of the face using the image generation network.

4. The method of claim 1 , further comprising:

fine-tuning the image generation network based on the pruned channels of the block of the encoder and the pruned channels of the block of the decoder.

5. The method of claim 1 , further comprising:

refraining from pruning a mapping network of the image generation network, wherein the encoder and the decoder are components of a synthesis network of the image generation network.

6. The method of claim 1 , further comprising:

refraining from pruning a global encoder block of the encoder and a global decoder block of the decoder.

7. The method of claim 1 , wherein:

the block of the encoder and the block of the decoder comprise convolutional layers.

8. A non-transitory computer readable medium storing code for image processing, the code comprising instructions that, when executed by at least one processor, cause the at least one processor to perform operations for an image generation method comprising:

identifying, by a machine learning model, an image generation network that includes an encoder and a decoder;

pruning channels of a first layer of a block of the encoder;

pruning channels of a second layer of the block of the encoder based on the pruned channels of the first layer of the block of the encoder;

pruning channels of a block of the decoder that is connected to the block of the encoder by a skip connection, wherein the channels of the block of the decoder are pruned based on the pruned channels of the block of the encoder;

wherein pruning channels of the block of the decoder comprises:

pruning channels of a first layer of the block of the decoder based on the pruned channels of the first layer of the block of the encoder; and

pruning channels of a second layer of the block of the decoder based on the pruned channels of the second layer of the block of the encoder; and

generating an image using the image generation network based on the pruned channels of the block of the encoder and the pruned channels of the block of the decoder.

9. The non-transitory computer readable medium of claim 8 , the code further comprising instructions executable by the at least one processor to perform operations comprising:

identifying an input image and a portion of the input image for inpainting; and

inpainting the portion of the input image using the image generation network to obtain an inpainted image.

10. The non-transitory computer readable medium of claim 8 , the code further comprising instructions executable by the at least one processor to perform operations comprising:

identifying an image of a face; and

generating an anonymized image of the face using the image generation network.

11. The non-transitory computer readable medium of claim 8 , the code further comprising instructions executable by the at least one processor to perform operations comprising:

fine-tuning the image generation network based on the pruned channels of the block of the encoder and the pruned channels of the block of the decoder.

12. The non-transitory computer readable medium of claim 8 , the code further comprising instructions executable by the at least one processor to perform operations comprising:

refraining from pruning a mapping network of the image generation network, wherein the encoder and the decoder are components of a synthesis network of the image generation network.

13. The non-transitory computer readable medium of claim 8 , the code further comprising instructions executable by the at least one processor to perform operations comprising:

refraining from pruning a global encoder block of the encoder and a global decoder block of the decoder.

14. A system for image generation comprising:

a memory component; and

a processing device coupled to the memory component, the processing device configured to perform operations comprising:

identifying, by a machine learning model, an image generation network that includes an encoder and a decoder;

pruning channels of a first layer of a block of the encoder;

pruning channels of a second layer of the block of the encoder based on the pruned channels of the first layer of the block of the encoder;

pruning channels of a block of the decoder that is connected to the block of the encoder by a skip connection, wherein the channels of the block of the decoder are pruned based on the pruned channels of the block of the encoder;

wherein pruning channels of the block of the decoder comprises:

pruning channels of a first layer of the block of the decoder based on the pruned channels of the first layer of the block of the encoder; and

pruning channels of a second layer of the block of the decoder based on the pruned channels of the second layer of the block of the encoder; and

generating an image using the image generation network based on the pruned channels of the block of the encoder and the pruned channels of the block of the decoder.

15. The system of claim 14 , wherein:

the image generation network includes a synthesis network and a mapping network, and wherein the synthesis network includes the encoder and the decoder.

16. The system of claim 14 , further comprising:

a decomposition component configured to perform tensor decomposition on a layer of the image generation network and to compress the layer of the image generation network based on the tensor decomposition.

Assignments (1)

Continuity (1)

References Cited (23)