公开(公告)号：US20230119791A1

公开(公告)日：2023-04-20

申请号：US17937765

申请日：2022-10-03

Applicant: QUALCOMM Incorporated

Inventor： Byeonggeun KIM ,  Seunghan YANG ,  Hyunsin PARK ,  Juntae LEE ,  Simyung CHANG

IPC: G10L21/034 ,  G10L17/18 ,  G10L25/30 ,  G10L25/51 ,  G10L17/04

Abstract: Techniques and apparatus for training a neural network to classify audio into one of a plurality of categories and using such a trained neural network. An example method generally includes receiving a data set including a plurality of audio samples. A relaxed feature-normalized data set is generated by normalizing each audio sample of the plurality of audio samples. A neural network is trained to classify audio into one of a plurality of categories based on the relaxed feature-normalized data set, and the trained neural network is deployed.

公开(公告)号：US20220405547A1

公开(公告)日：2022-12-22

申请号：US17807479

申请日：2022-06-17

Applicant: QUALCOMM Incorporated

Inventor： Byeonggeun KIM ,  Simyung Chang ,  Jangho Kim ,  Seunghan Yang ,  Kyu Woong Hwang

IPC: G06N3/04 ,  G06K9/62

Abstract: Certain aspects of the present disclosure provide techniques for residual normalization. A first tensor comprising a frequency dimension and a temporal dimension is accessed. A second tensor is generated by applying a frequency-based instance normalization operation to the first tensor, comprising, for each respective frequency bin in the frequency dimension, computing a respective frequency-specific mean of the first tensor. A third tensor is generated by: scaling the first tensor by a scale value, and aggregating the scaled first tensor and the second tensor. The third tensor is provided as input to a layer of a neural network.

公开(公告)号：US20220309344A1

公开(公告)日：2022-09-29

申请号：US17656621

申请日：2022-03-25

Applicant: QUALCOMM Incorporated

Inventor： Byeonggeun KIM ,  Simyung Chang ,  Jinkyu Lee ,  Dooyong Sung

IPC: G06N3/08 ,  G06F17/15

Abstract: Certain aspects of the present disclosure provide techniques for efficient broadcasted residual machine learning. An input tensor comprising a frequency dimension and a temporal dimension is received, and the input tensor is processed with a first convolution operation to generate a multidimensional intermediate feature map comprising the frequency dimension and the temporal dimension. The multidimensional intermediate feature map is converted to a one-dimensional intermediate feature map in the temporal dimension using a frequency dimension reduction operation, and the one-dimensional intermediate feature map is processed using a second convolution operation to generate a temporal feature map. The temporal feature map is expanded to the frequency dimension using a broadcasting operation to generate a multidimensional output feature map, and the multidimensional output feature map is augmented with the multidimensional intermediate feature map via a first residual connection.