# US 2025/0166653 A1 · Signal processing apparatus and signal processing method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0166653 A1
- Document ID: US20250166653A1
- Status: Published application
- Kind: A1
- Application number: 18/839,559
- Filed: 2023-01-17
- Published: 2025-05-22
- Earliest claimed: 2022-02-25
- Assignees: Sony Group Corporation
- Applicants: Sony Group Corporation
- Inventors: YOUSUKE TAKAHASHI; TOMOKI YAMASHIMA; MARIKO YOSHIOKA; TAKASHI KUDO; YOSUKE HORIBA
- Web page: https://patentgrep.com/patent/US20250166653A1

## CPC

- G10L 21/0364: for improving intelligibility
- H04R 3/04: for correcting frequency response
- G10L 21/0272: Voice signal separating
- G10L 21/034: Automatic adjustment
- G10L 25/30: using neural networks
- G10L 25/84: for discriminating voice from noise
- H04R 3/00: Circuits for transducers (arrangements for producing a reverberation or echo sound G10K15/08; amplifiers H03F)
- H04S 3/008: {in which the audio signals are in digital form, i.e. employing more than two discrete digital channels (data reduction aspects thereof based on psychoacoustics G10L19/02)}
- H04S 7/00: Indicating arrangements; Control arrangements, e.g. balance control
- G06F 3/165: {Management of the audio stream, e.g. setting of volume, audio stream path}
- H04R 2430/01: Aspects of volume control, not necessarily automatic, in sound systems
- H04S 2400/01: Multi-channel, i.e. more than two input channels, sound reproduction with two speakers wherein the multi-channel information is substantially preserved
- H04S 2400/13: Aspects of volume control, not necessarily automatic, in stereophonic sound systems
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- H04R: LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; ELECTRIC HEARING AIDS; PUBLIC ADDRESS SYSTEMS (producing sounds with frequency not determined by supply frequency G10K)
- H04S: STEREOPHONIC SYSTEMS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

To create, for example, an experience value of content. Provided is a signal processing apparatus including a feature extraction section that uses a learning model obtained through machine learning, to extract a signal of specific sound from an input signal, and a feature addition section that applies gain adjustment to the signal of the specific sound extracted in the feature extraction section and adds a result of the gain adjustment to a signal based on the input signal.

## Claims

### Claim 1

A signal processing apparatus comprising:
a feature extraction section that uses a learning model obtained through machine learning, to extract a signal of specific sound from an input signal; and
a feature addition section that applies gain adjustment to the signal of the specific sound extracted in the feature extraction section and adds a result of the gain adjustment to a signal based on the input signal.

### Claim 2 (depends on 1)

The signal processing apparatus according to claim 1, comprising:
a channel number conversion section that changes the number of channels of the input signal and outputs a result of the change,
wherein the feature addition section adds the signal of the specific sound to the signal output from the channel number conversion section.

### Claim 3 (depends on 2)

The signal processing apparatus according to claim 2,
wherein the channel number conversion section uses an upmixing technology to increase the number of channels.

### Claim 4 (depends on 1)

The signal processing apparatus according to claim 1,
wherein both the input signal and the signal based on the input signal are signals on multiple channels,
the feature extraction section extracts the signal of the specific sound from each channel signal of the input signals, and
the feature addition section applies the gain adjustment to each signal of the specific sound extracted in the feature extraction section and adds each result of the gain adjustment to each channel signal based on the input signal.

### Claim 5 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the feature addition section applies the gain adjustment to the signal of the specific sound with such a setting that clarity of the specific sound changes.

### Claim 6 (depends on 5)

The signal processing apparatus according to claim 5,
wherein the specific sound is a vocal of music content, and
the feature addition section reduces a gain as the gain adjustment.

### Claim 7 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the feature addition section applies the gain adjustment to the signal of the specific sound with such a setting that localization of the specific sound changes.

### Claim 8 (depends on 1)

The signal processing apparatus according to claim 1, comprising:
a delay processing section that delays output of the signal based on the input signal to the feature addition section according to a processing time in the feature extraction section.

### Claim 9 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the specific sound is sound relating to voice.

### Claim 10 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the specific sound is sound of a specific musical instrument, a sound effect, cheer sound, or noise.

### Claim 11 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the feature extraction section uses a DNN (Deep Neural Network) as the machine learning.

### Claim 12 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the feature addition section applies the gain adjustment to the signal of the specific sound with a predetermined fixed setting.

### Claim 13 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the feature addition section automatically applies the gain adjustment to the signal of the specific sound according to sensing information output from a sensor device.

### Claim 14 (depends on 13)

The signal processing apparatus according to claim 13,
wherein a camera is included in the sensor device, and
the feature addition section applies the gain adjustment to the signal of the specific sound according to a user age obtained by analyzing an image captured by the camera.

### Claim 15 (depends on 13)

The signal processing apparatus according to claim 13,
wherein a camera is included in the sensor device, and
the feature addition section applies the gain adjustment to the signal of the specific sound according to a user position obtained by analyzing an image captured by the camera.

### Claim 16 (depends on 13)

The signal processing apparatus according to claim 13,
wherein a microphone is included in the sensor device, and
the feature addition section applies the gain adjustment to the signal of the specific sound according to a level of external noise obtained by analyzing collected sound information obtained by the microphone.

### Claim 17 (depends on 1)

The signal processing apparatus according to claim 1,
wherein the feature addition section applies, as desired, the gain adjustment to the signal of the specific sound according to operation information output from a user interface.

### Claim 18

A signal processing method comprising:
a feature extraction step of using a learning model obtained through machine learning, to extract a signal of specific sound from an input signal; and
a feature addition step of applying gain adjustment to the signal of the specific sound extracted in the feature extraction step and adding a result of the gain adjustment to a signal based on the input signal.
