公开(公告)号：US20220375462A1

公开(公告)日：2022-11-24

申请号：US17874826

申请日：2022-07-27

Applicant: Samsung Electronics Co., Ltd.

Inventor： Sourav BHATTACHARYA ,  Abhinav MEHROTRA ,  Alberto Gil C. P. RAMOS

IPC: G10L15/16 ,  G10L21/0208

Abstract: Broadly speaking, the present techniques provide methods for conditioning a neural network, which not only improve the generalizable performance of conditional neural networks, but also reduce model size and latency significantly. The resulting conditioned neural network is suitable for on-device deployment due to having a significantly lower model size, lower dynamic memory requirement, and lower latency.

公开(公告)号：US20210081763A1

公开(公告)日：2021-03-18

申请号：US17015724

申请日：2020-09-09

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Mohamed S. ABDELFATTAH ,  Lukasz DUDZIAK ,  Chun Pong CHAU ,  Hyeji KIM ,  Royson LEE ,  Sourav BHATTACHARYA

IPC: G06N3/04 ,  G06N7/00 ,  G06N3/08

Abstract: Disclosed are an electronic device and a method for controlling thereof. The electronic device includes: a memory for storing a plurality of accelerators and a plurality of neural networks and a processor configured to: select a first neural network among the plurality of neural networks and select a first accelerator to implement the first neural network among the plurality of accelerators, implement the first neural network on the first accelerator to obtain information associated with the implementation, obtain a first reward value for the first accelerator and the first neural network based on the information associated with the implementation, select a second neural network to be implemented on the first accelerator among the plurality of neural networks, implement the second neural network on the first accelerator to obtain the information associated with the implementation, obtain a second reward value for the first accelerator and the second neural network based on the information associated with the implementation, and select a neural network and an accelerator having a largest reward value among the plurality of neural networks and the plurality of accelerators based on the first reward value and the second reward value.

公开(公告)号：US20230410818A1

公开(公告)日：2023-12-21

申请号：US18220567

申请日：2023-07-11

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Alberto Gil C. P. RAMOS ,  Abhinav MEHROTRA ,  Sourav BHATTACHARYA

IPC: G10L17/18 ,  G10L17/04 ,  G10L21/0208 ,  G10L13/08

CPC classification number: G10L17/18 ,  G10L13/08 ,  G10L21/0208 ,  G10L17/04

Abstract: Broadly speaking, embodiments of the present techniques provide a method and system for personalising machine learning models on resource-constrained devices by using conditional neural networks. In particular, the present techniques allow for resource-efficient use of a conditioning vector by incorporating the conditioning vector into weights learned during training. This reduces the computational resources required at inference time.

公开(公告)号：US20230298593A1

公开(公告)日：2023-09-21

申请号：US18322228

申请日：2023-05-23

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Alberto Gil RAMOS ,  Carlos PURVES ,  Abhinvav MEHROTRA ,  Sourav BHATTACHARYA ,  Ravichander VIPPERLA ,  Syed Samin ISHTIAQ ,  Nicholas Donald Atkins LANE

IPC: G10L17/20 ,  G10L17/04 ,  G10L17/18 ,  G10L21/0232 ,  G10L17/02

CPC classification number: G10L17/20 ,  G10L17/04 ,  G10L17/18 ,  G10L21/0232 ,  G10L17/02

Abstract: Broadly speaking, the present techniques generally relate to a system, computer-implemented method and apparatus for training a machine learning, ML, model to perform sound enhancement for a target user in real-time, and to a method and apparatus for using the trained ML model to perform sound enhancement of audio signals in real-time. Advantageously, the present techniques are suitable for implementation on resource-constrained devices that capture audio signals, such as smartphones and Internet of Things devices.

公开(公告)号：US20210350788A1

公开(公告)日：2021-11-11

申请号：US17198727

申请日：2021-03-11

Applicant: Samsung Electronics Co., Ltd.

Inventor： Kihyun CHOO ,  Sangjun PARK ,  Nicholas LANE ,  Ravichander VIPPERLA ,  Sourav BHATTACHARYA ,  Syed Samin ISHTIAQ ,  Taehwa KANG ,  Jonghoon JEONG

IPC: G10L13/08 ,  G10L13/02

Abstract: A method, performed by an electronic device, of generating a speech signal corresponding to at least one text is provided. The method includes obtaining feature information with respect to a first sample included in the speech signal, based on the at least one text, obtaining condition information related to a condition under which a bunching operation, in which one or more sample values included in the speech signal are obtained, is performed, based on the feature information, configuring one or more bunching blocks for performing the bunching operation, based on the condition information, obtaining the one or more sample values based on the feature information with respect to the first sample by using the one or more bunching blocks, and generating the speech signal based on the obtained one or more sample values.