公开(公告)号：US20230088157A1

公开(公告)日：2023-03-23

申请号：US17794537

申请日：2020-01-28

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Shoichiro SAITO

IPC分类号： G01M99/00 ,  G06N3/08

摘要： An anomaly degree calculation device 200 includes an anomaly degree calculation unit 201 that calculates an anomaly degree on a basis of a feature amount extracted from target data that is a calculation target of the anomaly degree. The anomaly degree calculation unit 201 calculates the anomaly degree on a basis of a similarity degree of the target data and registration data registered in advance. The similarity degree is calculated in consideration of a degree to which a frame constituting the target data and a frame constituting the registration data are similar to each other.

公开(公告)号：US20210229604A1

公开(公告)日：2021-07-29

申请号：US17263867

申请日：2019-07-10

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Shoichiro SAITO ,  Masahiro FUKUI

IPC分类号： B60R11/02 ,  H04R3/00 ,  H04R1/40

摘要： An object is to provide a sound collection apparatus capable of collecting a sound with a higher SN rate than before through use of a reflected sound. The sound collection apparatus is installed in a vehicle. The vehicle includes a seat for a passenger to sit on, and a reflective structure that reflects a sound emitted by a passenger sitting in the seat. The sound collection apparatus includes a first microphone arranged at a position where a direct sound that is a sound emitted by the passenger is readily collected, and a second microphone arranged at a position where a reflected sound that is a sound emitted by the passenger and reflected by the reflective structure is readily collected.

公开(公告)号：US20190120719A1

公开(公告)日：2019-04-25

申请号：US16089214

申请日：2017-03-31

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Shoichiro SAITO ,  Hisashi UEMATSU ,  Kenta NIWA ,  Hiroaki ITO

IPC分类号： G01M13/028 ,  G06N3/08 ,  G05B19/4065

摘要： An anomalous sound detection training apparatus includes: a first acoustic feature extraction unit that extracts an acoustic feature of normal sound based on training data for normal sound by using an acoustic feature extractor; a normal sound model updating unit that updates a normal sound model by using the acoustic feature extracted; a second acoustic feature extraction unit that extracts an acoustic feature of anomalous sound based on simulated anomalous sound and extracts the acoustic feature of normal sound based on the training data for normal sound by using the acoustic feature extractor; and an acoustic feature extractor updating unit that updates the acoustic feature extractor by using the acoustic feature of anomalous sound and the acoustic feature of normal sound that have been extracted, in which processing by the units is repeatedly performed.

公开(公告)号：US20220189499A1

公开(公告)日：2022-06-16

申请号：US17600029

申请日：2020-03-23

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Shoichiro SAITO ,  Hiroaki ITO

IPC分类号： G10L21/034 ,  G10L15/22 ,  G10L25/51

摘要： Provided are a volume control apparatus capable of appropriately controlling a sound volume even immediately after start of utterance, an associated method, and a program. The volume control apparatus includes a recognition unit that recognizes a predetermined voice command for use in starting voice recognition, a gain setting unit that sets a gain for an audio signal X of a target of the voice recognition, by use of an audio signal related to the predetermined voice command uttered by a user, and an adjustment unit that adjusts a sound volume of the audio signal X, by use of the gain.

公开(公告)号：US20200209842A1

公开(公告)日：2020-07-02

申请号：US16643916

申请日：2018-08-24

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Yuta KAWACHI ,  Noboru HARADA ,  Shoichiro SAITO ,  Akira NAKAGAWA ,  Shin MURATA

IPC分类号： G05B23/02 ,  G06N20/00

摘要： Accuracy of unsupervised anomalous sound detection is improved using a small number of pieces of anomalous sound data. A threshold deciding part (13) calculates an anomaly score for each of a plurality of pieces of anomalous sound data, using a normal model learned with normal sound data and an anomaly model expressing the pieces of anomalous sound data, and decides a minimum value among the anomaly scores as a threshold. A weight updating part (14) updates, using a plurality of pieces of normal sound data, the pieces of anomalous sound data and the threshold, weights of the anomaly model so that all the pieces of anomalous sound data are judged as anomalous, and probability of the pieces of normal sound data being judged as anomalous is minimized.

公开(公告)号：US20230086628A1

公开(公告)日：2023-03-23

申请号：US17798849

申请日：2020-02-12

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Shoichiro SAITO ,  Hisashi UEMATSU ,  Shin MURATA

IPC分类号： G06K9/62 ,  G06N3/04

摘要： Provided is an abnormal data generation device capable of generating highly accurate abnormal data. The abnormal data generation device includes an abnormal data generation unit for generating pseudo generated data of abnormal data that has, in the same latent space, a normal distribution as a normal data generation model and an abnormal distribution expressed as a complementary set of the normal distribution and that is optimized such that pseudo generated data cannot be discriminated from observed actual abnormal data by a latent variable sampled from the abnormal distribution.

公开(公告)号：US20220199083A1

公开(公告)日：2022-06-23

申请号：US17603263

申请日：2020-04-03

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Shoichiro SAITO ,  Hiroaki ITO

IPC分类号： G10L15/22 ,  G10L15/05 ,  G10L25/87

摘要： A command analysis device capable of shortening the time until a command is executed is provided. The command analysis device includes a speech recognition unit that performs, every time when a predetermined unit of a speech signal is input, speech recognition on the speech signal and acquires a partial speech recognition result which is an intermediate result, and a command analysis unit that verifies an intermediate result in light of a predetermined intermediate result recognition rule, and outputs an analysis result during the input of the speech signal when a command execution target and a command execution content are successfully analyzed are provided.

公开(公告)号：US20210321192A1

公开(公告)日：2021-10-14

申请号：US17265673

申请日：2019-07-25

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Sakurako YAZAWA ,  Kazunori KOBAYASHI ,  Kenichi NOGUCHI ,  Shoichiro SAITO

IPC分类号： H04R1/40

摘要： Provided is a sound collection apparatus, including: a substantially spherical base on which at least N recess portions are provided in a surface thereof with a predetermined interval therebetween; and N microphones. N is an integer of 2 or greater. The microphones are installed on an inner bottom surface side of the recess portions one by one. Intervals between sound collecting portions of adjacent microphones are substantially equal to each other.

公开(公告)号：US20210235191A1

公开(公告)日：2021-07-29

申请号：US17263860

申请日：2019-07-10

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Shoichiro SAITO ,  Kazunori KOBAYASHI

IPC分类号： H04R3/12 ,  H04R1/40 ,  H04R1/02 ,  H04R3/00 ,  H04R3/02 ,  H03G3/30

摘要： Provided is a sound pickup loudspeaker apparatus which suppresses the influence of feedback produced when adding a speaker while also reducing a sense of unnaturalness in a sound image. The sound pickup loudspeaker apparatus collects a target sound emitted from a first seat in the vehicle, plays back the collected target sound to a listener seated in a second seat in the vehicle from a second sound amplifying device disposed in a direction different from the first seat from the perspective of the listener, and furthermore plays back the collected target sound to the listener from a first sound amplifying device disposed in the same direction as the first seat from the perspective of the listener. The apparatus multiplies the target sound by a first gain and outputs the target sound to the first sound amplifying device; delays the target sound by an amount of time obtained by adding a time for achieving a precedence effect to a delay time of the first sound amplifying device relative to the second sound amplifying device; and multiplies the delayed target sound by a second gain and outputs the target sound to the second sound amplifying device. A first gain adjustment unit adjusts the first gain to a low value and a second gain adjustment unit adjusts the second gain to a high value.

公开(公告)号：US20210233552A1

公开(公告)日：2021-07-29

申请号：US17263522

申请日：2019-07-10

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Shoichiro SAITO

IPC分类号： G10L21/0324 ,  H04R3/00

摘要： Provided is a sound pickup loudspeaker apparatus that has a mechanism for detecting situations in which sound amplification should not be performed and stopping sound amplification from a speaker. The sound pickup loudspeaker apparatus is installed in a vehicle. At least a sound pickup position and a sound amplification position are assumed to be present in the vehicle. The sound pickup loudspeaker apparatus includes a noise level estimation unit that determines a noise level that is an estimated value of the magnitude of a noise component included in a sound pickup signal obtained from a microphone that collects sound emitted from the sound pickup position, and a control unit that performs control such that (i) when the noise level is not greater than a predetermined threshold value or is less than the threshold value, an acoustic signal based on the sound pickup signal is played back from a speaker at the sound amplification position, and (ii) when the noise level is greater than the threshold value or is at least the threshold value, the acoustic signal is not played back from the speaker.