公开(公告)号：US20230154171A1

公开(公告)日：2023-05-18

申请号：US17720681

申请日：2022-04-14

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Seohyung LEE ,  Dongwook LEE ,  Changbeom PARK ,  Byung In YOO

IPC: G06V10/82 ,  G06V10/40 ,  G06N3/08

CPC classification number: G06V10/82 ,  G06N3/08 ,  G06V10/40

Abstract: A method with self-attention includes: obtaining a three-dimensional (3D) feature map; generating 3D query data and 3D key data by performing a convolution operation based on the 3D feature map; generating two-dimensional (2D) vertical data based on a vertical projection of the 3D query data and the 3D key data; generating 2D horizontal data based on a horizontal projection of the 3D query data and the 3D key data; determining an intermediate attention result through a multiplication based on the 2D vertical data and the 2D horizontal data; and determining a final attention result through a multiplication based on the intermediate attention result and the 3D feature map.

公开(公告)号：US20210279568A1

公开(公告)日：2021-09-09

申请号：US17015122

申请日：2020-09-09

Applicant: Samsung Electronics Co., Ltd.

Inventor： Jaehyoung YOO ,  Jinwoo SON ,  Changyong SON ,  Seohyung LEE ,  Sangil JUNG ,  Changin CHOI

IPC: G06N3/08 ,  G06N3/04 ,  G06F9/38

Abstract: Disclosed are methods and apparatuses for processing a convolution operation on a layer in a neural network. The method includes extracting a first target feature vector from a target feature map, extracting a first weight vector matched with the first target feature vector from a first-type weight element, based on matching relationships for depth-wise convolution operations between target feature vectors of the target feature map and weight vectors of the first-type weight element, generating a first intermediate feature vector by performing multiplication between the first target feature vector and the first weight vector, generating a first hidden feature vector by accumulating the first intermediate feature vector and a second intermediate feature vector generated based on a second target feature vector, and generating a first output feature vector of an output feature map based on a point-wise convolution operation between the first hidden feature vector and a second-type weight element.

公开(公告)号：US20210142041A1

公开(公告)日：2021-05-13

申请号：US17075164

申请日：2020-10-20

Applicant: Samsung Electronics Co., Ltd.

Inventor： Seohyung LEE ,  Jaehyoung YOO ,  Jinwoo SON ,  Changyong SON ,  Sangil JUNG ,  Changin CHOI

IPC: G06K9/00 ,  G06T7/00 ,  G06T3/40

Abstract: Disclosed is a method and apparatus for face detection using an adaptive threshold. The method includes determining a detection box in an input image, calculating a confidence score indicating whether an object in the detection box corresponds to a face, setting an adaptive threshold based on a size of the detection box, and determining whether the object in the detection box corresponds to a face based on comparing the confidence score to the adaptive threshold.

公开(公告)号：US20200074058A1

公开(公告)日：2020-03-05

申请号：US16527332

申请日：2019-07-31

Applicant: Samsung Electronics Co., Ltd.

Inventor： Jinwoo SON ,  Changyong SON ,  Jaejoon HAN ,  Sangil JUNG ,  Seohyung LEE

IPC: G06F21/32 ,  G06N3/08 ,  G06N3/04

Abstract: Disclosed is a method and apparatus for training a user terminal. A user terminal may authenticate a user input using an authentication model of the user terminal, generate a gradient to train the authentication model from the user input, in response to a success in the authentication, accumulate the generated gradient in positive gradients, and train the authentication model based on the positive gradients.

公开(公告)号：US20240420443A1

公开(公告)日：2024-12-19

申请号：US18818219

申请日：2024-08-28

Applicant: Samsung Electronics Co., Ltd.

Inventor： Sangil JUNG ,  Dongwook LEE ,  Jinwoo SON ,  Changyong SON ,  Jaehyoung YOO ,  Seohyung LEE ,  Changin CHOI ,  Jaejoon HAN

IPC: G06V10/28 ,  G06F18/214 ,  G06F18/24 ,  G06T3/4046 ,  G06V20/10

Abstract: A system includes: an image sensor configured to acquire an image; an image processor configured to generate a quantized image based on the acquired image using a trained quantization filter; and an output interface configured to output the quantized image.

公开(公告)号：US20230229926A1

公开(公告)日：2023-07-20

申请号：US18127891

申请日：2023-03-29

Applicant: Samsung Electronics Co., Ltd.

Inventor： Chang Kyu CHOI ,  Youngjun KWAK ,  Seohyung LEE

IPC: G06N3/084 ,  G06T5/00 ,  G06T5/20 ,  G06N3/08 ,  G06N20/00 ,  G06V10/764 ,  G06V10/82 ,  G06V10/44

CPC classification number: G06N3/084 ,  G06T5/001 ,  G06T5/20 ,  G06N3/08 ,  G06N20/00 ,  G06V10/764 ,  G06V10/82 ,  G06V10/451 ,  G06T2207/10024 ,  G06T2207/20084 ,  G06F2218/00

Abstract: A processor-implemented method of generating feature data includes: receiving an input image; generating, based on a pixel value of the input image, at least one low-bit image having a number of bits per pixel lower than a number of bits per pixel of the input image; and generating, using at least one neural network, feature data corresponding to the input image from the at least one low-bit image.

公开(公告)号：US20210049474A1

公开(公告)日：2021-02-18

申请号：US16856112

申请日：2020-04-23

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Jinwoo SON ,  Changyong SON ,  Jaehyoung YOO ,  Seohyung LEE ,  Sangil JUNG ,  Changin CHOI ,  Jaejoon HAN

IPC: G06N3/08

Abstract: A processor-implemented data processing method and apparatus for a neural network is provided. The data processing method includes generating cumulative data by accumulating results of multiplication operations between at least a portion of input elements in an input plane and at least a portion of weight elements in a weight plane, and generating an output plane corresponding to an output channel among output planes of an output feature map respectively corresponding to output channels based on the generated cumulative data.

公开(公告)号：US20190213400A1

公开(公告)日：2019-07-11

申请号：US16182861

申请日：2018-11-07

Applicant: Samsung Electronics Co., Ltd.

Inventor： Youngsung KIM ,  Youngjun KWAK ,  Byung In YOO ,  Seohyung LEE

IPC: G06K9/00 ,  G06K9/62 ,  G10L15/06 ,  G10L15/02

CPC classification number: G06K9/00315 ,  G06K9/00335 ,  G06K9/00597 ,  G06K9/6262 ,  G10L15/02 ,  G10L15/063

Abstract: A method and apparatus with emotion recognition acquires a plurality of pieces of data corresponding a plurality of inputs for each modality and corresponding to a plurality of modalities; determines a dynamics representation vector corresponding to each of the plurality of modalities based on a plurality of features for each modality extracted from the plurality of pieces of data; determines a fused representation vector based on the plurality of dynamics representation vectors corresponding to the plurality of modalities; and recognizes an emotion of a user based on the fused representation vector.

公开(公告)号：US20190213399A1

公开(公告)日：2019-07-11

申请号：US16225084

申请日：2018-12-19

Applicant: Samsung Electronics Co., Ltd.

Inventor： Byung In YOO ,  Youngjun KWAK ,  Youngsung KIM ,  Seohyung LEE

IPC: G06K9/00 ,  G10L25/63 ,  G10L17/02

CPC classification number: G06K9/00302 ,  G06K9/00281 ,  G06K9/00315 ,  G06K9/6274 ,  G06N3/04 ,  G06N3/084 ,  G10L25/78

Abstract: A facial expression recognition apparatus and method and a facial expression training apparatus and method are provided. The facial expression recognition apparatus generates a speech map indicating a correlation between a speech and each portion of an object based on a speech model, extracts a facial expression feature associated with a facial expression based on a facial expression model, and recognizes a facial expression of the object based on the speech map and the facial expression feature. The facial expression training apparatus trains the speech model and the facial expression model.