公开(公告)号：US20250029217A1

公开(公告)日：2025-01-23

申请号：US18527955

申请日：2023-12-04

Applicant: Samsung Electronics Co., Ltd.

Inventor： Joonsoo Kim ,  Tien C. Bau

IPC: G06T5/00

Abstract: A method includes receiving a degraded image including aliasing artifacts and inputting the degraded image to an image enhancement network. This method also includes processing, using the image enhancement network, the degraded image to remove one or more of the aliasing artifacts and outputting, by the image enhancement network, a restored high-quality image. Another method includes obtaining a high-quality image of an environment and generating at least one degraded image of the environment by performing an aliasing artifact simulation on the obtained high-quality image. Performing the aliasing artifact simulation includes (i) performing a broken line artifact simulation to introduce one or more broken line artifacts on one or more objects in the environment of the high-quality image and/or (ii) performing a jaggy artifact simulation to introduce jaggy edges to one or more other objects in the environment of the high-quality image.

公开(公告)号：US20240412556A1

公开(公告)日：2024-12-12

申请号：US18527036

申请日：2023-12-01

Applicant: Samsung Electronics Co., Ltd.

Inventor： Sumit Jha ,  Tien C. Bau

IPC: G06V40/16 ,  G06V10/70

Abstract: A method for using a machine learning model having a base model and multiple branch models includes obtaining an image containing a face and processing the image using the base model to generate intermediate data. The method also includes processing the intermediate data using a first of the branch models to perform a first image processing task, where the first image processing task is associated with analyzing the image containing the face. The method further includes processing the intermediate data using a second of the branch models to perform a second image processing task different from the first image processing task, where the second image processing task is associated with analyzing the image containing the face. The base model and the first branch model are trained using a first dataset, and the base model and the second branch model are trained using a second dataset different from the first dataset.

公开(公告)号：US20230153625A1

公开(公告)日：2023-05-18

申请号：US18052297

申请日：2022-11-03

Applicant: Samsung Electronics Co., Ltd.

Inventor： Tien C. Bau ,  Arshita Gupta ,  Hrishikesh Deepak Garud

IPC: G06N3/082

CPC classification number: G06N3/082

Abstract: A method includes accessing a machine learning model, the machine learning model trained using a torque-based constraint. The method also includes receiving an input from an input source and providing the input to the machine learning model. The method also includes receiving an output from the machine learning model. The method also includes instructing at least one action based on the output from the machine learning model. Training the machine learning model includes applying a torque-based constraint on one or more filters of the machine learning model, adjusting, based on applying the torque-based constraint, a first set of one or more filters of the machine learning model to have a higher concentration of weights than a second set of one or more filters of the machine learning model, and pruning at least one channel of the machine learning model based on an average weight for the at least one channel.

公开(公告)号：US20250029210A1

公开(公告)日：2025-01-23

申请号：US18528241

申请日：2023-12-04

Applicant: Samsung Electronics Co., Ltd.

Inventor： Joonsoo Kim ,  Tien C. Bau ,  Kamal Jnawali

IPC: G06T5/50 ,  G06T3/4053 ,  G06T5/40

Abstract: A method includes obtaining a first image and a second image, where the second image represents a super-resolution version of the first image. The method also includes generating a third image representing a higher-resolution version of the first image. The method further includes performing flicker detection based on the third image and the second image in order to identify one or more flicker regions. The method also includes performing frequency decomposition of the third image to generate first decomposed images and of the second image to generate second decomposed images. In addition, the method includes blending portions of at least some of the first and second decomposed images based on the one or more identified flicker regions to generate a flicker-suppressed image.

公开(公告)号：US20230081128A1

公开(公告)日：2023-03-16

申请号：US17879647

申请日：2022-08-02

Applicant: Samsung Electronics Co., Ltd.

Inventor： Tien C. Bau ,  Hrishikesh Deepak Garud

IPC: G06T7/11 ,  G06T3/40

Abstract: A method includes training a semantic segmentation network to generate semantic segmentation maps having class-wise probability values. The method also includes generating a semantic segmentation map using the trained semantic segmentation network. The method further includes utilizing the semantic segmentation map during training of an image generation network as part of a loss function that includes multiple losses. The semantic segmentation network may be trained to be sensitive to picture quality of an output image generated by the image generation network during the training of the image generation network such that increased degradation of the picture quality of the output image results in decreased prediction confidence by the semantic segmentation network. The semantic segmentation network may be trained to vary the class-wise probability values based on the picture quality.

公开(公告)号：US20220121937A1

公开(公告)日：2022-04-21

申请号：US17443666

申请日：2021-07-27

Applicant: Samsung Electronics Co., Ltd.

Inventor： Chang Su ,  Tien C. Bau ,  Arthita Ghosh

IPC: G06N3/08 ,  G06N3/04 ,  G06N3/10

Abstract: A method includes processing, using at least one processor of an electronic device, input data using a first layer of a neural network to generate a feature map. The method also includes representing, using the at least one processor, feature data of the feature map using index values. The index values correspond to multiple records of a look up table (LUT), and the records of the LUT represent a non-uniform distribution of quantization levels of the feature map. The method further includes storing, using the at least one processor, the index values in a memory of the electronic device. The method also includes regenerating, using the at least one processor, the feature data of the feature map by cross-referencing the index values with the LUT. In addition, the method includes processing, using the at least one processor, the feature data using a second layer of the neural network.