公开(公告)号：US20230334318A1

公开(公告)日：2023-10-19

申请号：US18341050

申请日：2023-06-26

Applicant: Samsung Electronics Co., Ltd.

Inventor： Qingfeng LIU ,  Mostafa EL-KHAMY ,  Jungwon LEE ,  Behnam Babagholami MOHAMADABADI

IPC: G06N3/08 ,  G06F18/2323 ,  G06F18/2415 ,  G06F18/2431 ,  G06V30/19 ,  G06V10/772 ,  G06V10/774 ,  G06V10/80 ,  G06F18/2321 ,  G06F18/25 ,  G06N3/045 ,  G06N3/04 ,  G06N5/022 ,  G06N5/025

CPC classification number: G06N3/08 ,  G06F18/2323 ,  G06F18/2415 ,  G06F18/2431 ,  G06V30/19107 ,  G06V30/1914 ,  G06V30/19147 ,  G06V30/1918 ,  G06V10/772 ,  G06V10/774 ,  G06V10/806 ,  G06V10/809 ,  G06F18/2321 ,  G06F18/253 ,  G06F18/254 ,  G06N3/045 ,  G06N3/04 ,  G06N5/022 ,  G06N5/025

Abstract: A method and system for training a neural network are provided. The method includes receiving an input image, selecting at least one data augmentation method from a pool of data augmentation methods, generating an augmented image by applying the selected at least one data augmentation method to the input image, and generating a mixed image from the input image and the augmented image.

公开(公告)号：US20220300819A1

公开(公告)日：2022-09-22

申请号：US17826606

申请日：2022-05-27

Applicant: Samsung Electronics Co., Ltd.

Inventor： Haoyu REN ,  Mostafa EL-KHAMY ,  Jungwon LEE

IPC: G06N3/08 ,  G06T5/00 ,  G06T3/40 ,  G06N3/04

Abstract: Apparatuses and methods of manufacturing same, systems, and methods are described. In one aspect, a method includes generating a convolutional neural network (CNN) by training a CNN having a plurality of convolutional layers, and performing cascade training on the trained CNN. The cascade training includes an iterative process of a plurality of stages, in which each stage includes inserting a residual block (ResBlock) and training the CNN with the inserted ResBlock.

公开(公告)号：US20200304147A1

公开(公告)日：2020-09-24

申请号：US16576166

申请日：2019-09-19

Applicant: Samsung Electronics Co., Ltd.

Inventor： Yoo Jin CHOI ,  Mostafa EL-KHAMY ,  Jungwon LEE

IPC: H03M7/30 ,  H04N19/13 ,  H04N19/147 ,  H04N19/19 ,  H04L1/00 ,  G06N3/04 ,  G06N3/08

Abstract: A method and apparatus for variable rate compression with a conditional autoencoder is herein provided. According to one embodiment, a method includes training a conditional autoencoder using a Lagrange multiplier and training a neural network that includes the conditional autoencoder with mixed quantization bin sizes.

公开(公告)号：US20200227070A1

公开(公告)日：2020-07-16

申请号：US16451969

申请日：2019-06-25

Applicant: Samsung Electronics Co., Ltd.

Inventor： Jaeyoung KIM ,  Mostafa EL-KHAMY ,  Jungwon LEE

IPC: G10L25/60 ,  G10L15/06 ,  G10L21/0208

Abstract: A method and system for providing end-to-end multi-task denoising for joint signal distortion ratio (SDR) and perceptual evaluation of speech quality (PESQ) optimization is herein disclosed. According to one embodiment, an method includes receiving a noisy signal, generating a denoised output signal, determining a signal distortion ratio (SDR) loss function based on the denoised output signal, determining a perceptual evaluation of speech quality (PESQ) loss function based on the denoised output signal, and optimizing an overall loss function based on the PESQ loss function and the SDR loss function.

公开(公告)号：US20180268284A1

公开(公告)日：2018-09-20

申请号：US15655557

申请日：2017-07-20

Applicant: Samsung Electronics Co., Ltd.

Inventor： Haoyu REN ,  Mostafa EL-KHAMY ,  Jungwon LEE

IPC: G06N3/04 ,  G06N3/08

CPC classification number: G06N3/04 ,  G06N3/0454 ,  G06N3/08 ,  G06N3/082

Abstract: Apparatuses and methods of manufacturing same, systems, and methods for generating a convolutional neural network (CNN) are described. In one aspect, a minimal CNN having, e.g., three or more layers is trained. Cascade training may be performed on the trained CNN to insert one or more intermediate layers until a training error is less than a threshold. When cascade training is complete, cascade network trimming of the CNN output from the cascade training may be performed to improve computational efficiency. To further reduce network parameters, convolutional filters may be replaced with dilated convolutional filters with the same receptive field, followed by additional training/fine-tuning.

公开(公告)号：US20180060720A1

公开(公告)日：2018-03-01

申请号：US15343882

申请日：2016-11-04

Applicant: Samsung Electronics Co., Ltd.

Inventor： Mostafa EL-KHAMY ,  Jungwon LEE ,  Jaeyoung KIM

IPC: G06N3/04 ,  G06N3/08

CPC classification number: G06N3/04 ,  G06N3/0445 ,  G06N3/0454 ,  G06N3/063 ,  G06N3/08

Abstract: An apparatus and a method. The apparatus includes a first recurrent network in a first layer; a second recurrent network in a second layer connected to the first recurrent network; a distant input gate connected to the second recurrent network; a first highway gate connected to the distant input gate and the second recurrent network; a first elementwise product projection gate connected to the distant input gate, the highway gate, and the second recurrent network; a second highway gate connected to the first recurrent network and the second recurrent network; and a second elementwise product projection gate connected to the first recurrent network, the second highway gate, and the second recurrent network.

公开(公告)号：US20250053811A1

公开(公告)日：2025-02-13

申请号：US18639937

申请日：2024-04-18

Applicant: Samsung Electronics Co., Ltd.

Inventor： Ahmed ELKORDY ,  Behnam BABAGHOLAMI MOHAMADABADI ,  Mostafa EL-KHAMY

IPC: G06N3/082 ,  G06N3/04

Abstract: A system and a method are disclosed for hardware-aware pruning of conformer networks. In some embodiments, the method includes: training a neural network, the training including: performing a first pruning operation, on the neural network, after a first training epoch, and performing a second pruning operation, on the neural network, after a second training epoch and after the first pruning operation, wherein each of the pruning operations results in a respective pruning fraction, the respective pruning fraction being a function of an index of a training epoch preceding the pruning operation.

公开(公告)号：US20240177318A1

公开(公告)日：2024-05-30

申请号：US18107173

申请日：2023-02-08

Applicant: Samsung Electronics Co., Ltd.

Inventor： Mostafa EL-KHAMY ,  Hai SU ,  Qingfeng LIU

IPC: G06T7/12

CPC classification number: G06T7/12 ,  G06T2207/10016 ,  G06T2207/20081

Abstract: Disclosed is a method including receiving, in a semantic segmentation network, input data from a plurality of frames, computing a ground truth label on the plurality of frames, generating a ground truth temporal semantic boundary map from the ground truth label on the plurality of frames, generating a predicted temporal semantic boundary map based on an output of the input data, and determining a loss based on the ground truth temporal semantic boundary map and the predicted temporal semantic boundary map.

公开(公告)号：US20230050573A1

公开(公告)日：2023-02-16

申请号：US17825391

申请日：2022-05-26

Applicant: Samsung Electronics Co., Ltd.

Inventor： Qingfeng LIU ,  Mostafa EL-KHAMY ,  Yuewei YANG

IPC: G06N20/00 ,  G06V10/762 ,  G06V10/77

Abstract: Apparatuses and methods are provided for training a feature extraction model determining a loss function for use in unsupervised image segmentation. A method includes determining a clustering loss from an image; determining a weakly supervised contrastive loss of the image using cluster pseudo labels based on the clustering loss; and determining the loss function based on the clustering loss and the weakly supervised contrastive loss.

公开(公告)号：US20220083861A1

公开(公告)日：2022-03-17

申请号：US17532323

申请日：2021-11-22

Applicant: Samsung Electronics Co., Ltd.

Inventor： Xianzhi DU ,  Mostafa EL-KHAMY ,  Jungwon LEE

IPC: G06N3/08 ,  G06N20/00 ,  G06K9/62 ,  G06N3/04 ,  G06N7/00

Abstract: Methods and apparatuses for deep learning training are provided which include receiving a candidate unit for classification. The candidate unit including an intersection area between a ground-truth bounding box and a detection box. The candidate unit is classified by assigning a label that is a probability value that a given feature is observed in the intersection area. Deep learning training is performed using the assigned label of the classified candidate unit.