公开(公告)号：US11494612B2

公开(公告)日：2022-11-08

申请号：US16176812

申请日：2018-10-31

Applicant: Sony Interactive Entertainment Inc.

Inventor： Ruxin Chen ,  Min-Hung Chen ,  Jaekwon Yoo ,  Xiaoyu Liu

IPC: G06N3/04 ,  G06N3/08

Abstract: A domain adaptation module is used to optimize a first domain derived from a second domain using respective outputs from respective parallel hidden layers of the domains.

公开(公告)号：US20220067384A1

公开(公告)日：2022-03-03

申请号：US17105375

申请日：2020-11-25

Applicant: Sony Interactive Entertainment Inc.

Inventor： Lakshmish Kaushik ,  Saket Kumar ,  Jaekwon Yoo ,  Kevin Zhang ,  Soheil Khorram ,  Sharath Rao ,  Chockalingam Ravi Sundaram

IPC: G06K9/00 ,  G06N20/00

Abstract: Video and audio from a computer simulation are processed by a machine learning engine to identify candidate segments of the simulation for use in a video summary of the simulation. Text input is then used to reinforce whether a candidate segment should be included in the video summary.

公开(公告)号：US20200267473A1

公开(公告)日：2020-08-20

申请号：US16279896

申请日：2019-02-19

Applicant: Sony Interactive Entertainment Inc.

Inventor： Takeo Matsukawa ,  Jaekwon Yoo

IPC: H04R3/04 ,  H04R1/22 ,  H04R1/32 ,  H04R13/00 ,  H03F3/181

Abstract: An audio converter system is provided. The system comprises an audio input configured to receive a source audio, an audio output configured to couple to a hybrid speaker comprising at least two nondirectional speakers and a directional speaker, and a processor configured to generate an output audio for the hybrid speaker based on the source audio by: identifying a specific sound in the source audio, isolating the specific sound from the source audio, generating a directional speaker output for the directional speaker of the hybrid speaker based on the specific sound, and generating at least two channels of nondirectional speaker output for the at least two nondirectional speakers of the hybrid speaker.

公开(公告)号：US20190341025A1

公开(公告)日：2019-11-07

申请号：US16383896

申请日：2019-04-15

Applicant: Sony Interactive Entertainment Inc.

Inventor： Masanori Omote ,  Ruxin Chen ,  Xavier Menendez-Pidal ,  Jaekwon Yoo ,  Koji Tashiro ,  Sudha Krishnamurthy ,  Komath Naveen Kumar

IPC: G10L15/16 ,  G10L15/26

Abstract: A system and method for multimodal classification of user characteristics is described. The method comprises receiving audio and other inputs, extracting fundamental frequency information from the audio input, extracting other feature information from the video input, classifying the fundamental frequency information, textual information and video feature information using the multimodal neural network.