# US 2019/0294954 A1 · Artificial intelligence server

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

- Number: US 2019/0294954 A1
- Document ID: US20190294954A1
- Status: Published application
- Kind: A1
- Application number: 16/361,032
- Filed: 2019-03-21
- Published: 2019-09-26
- Earliest claimed: 2018-03-21
- Assignees: LG Electronics Inc.
- Applicants: LG Electronics Inc.
- Inventors: Jongwoo HAN; Hangil JEONG
- Web page: https://patentgrep.com/patent/US20190294954A1

## CPC

- G06N 3/08: Learning methods
- G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade}
- G06F 18/251: {of input or preprocessed data}
- G06N 3/045: Combinations of networks
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 3/0475: Generative networks
- G06N 3/09: Supervised learning
- G06N 3/094: Adversarial learning
- G06V 10/82: using neural networks
- G06V 20/10: Terrestrial scenes (scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59)
- G06V 20/20: in augmented reality scenes
- G10L 15/16: using artificial neural networks
- G10L 15/00: Speech recognition (G10L17/00 takes precedence)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING

## Abstract

Disclosed is an artificial intelligence (AI) server. The AI server includes a communication unit configured to communicate with an AI device; and an AI unit configured to receive feature data from the AI device, wherein the received feature data is generated by the AI device by obtaining sensing data and compressing the sensing data while preserving a feature of the sensing data; and input the received feature data to a deep learning model to obtain second sensing data for use in a recognition model related to an AI function of the AI device.

## Claims

### Claim 1

An artificial intelligence (AI) terminal comprising:
a communication unit configured to communicate with an AI device; and
an AI unit configured to:
receive feature data from the AI device, wherein the received feature data is generated by the AI device by obtaining sensing data and compressing the sensing data while preserving a feature of the sensing data; and
input the received feature data to a deep learning model to obtain second sensing data for use in a recognition model related to an AI function of the AI device.

### Claim 2 (depends on 1)

The AI terminal of claim 1, wherein the second sensing data includes a representation of the feature along with the compressed sensing data.

### Claim 3 (depends on 1)

The AI terminal of claim 1, wherein the AI unit is further configured use the second sensing data to train the recognition model for performing the AI function and transmit the trained recognition model to the AI device.

### Claim 4 (depends on 3)

The AI terminal of claim 3, wherein the feature data is received from the AI device when the AI device fails to perform the AI function using the sensing data.

### Claim 5 (depends on 1)

The AI terminal of claim 1, wherein:
an encoder of a variational auto-encoder is used to compress the sensing data, and the deep learning model is a decoder of the variational auto-encoder.

### Claim 6 (depends on 1)

The AI terminal of claim 1, wherein the deep learning model is a generator of a generative adversarial network (GAN).

### Claim 7 (depends on 1)

The AI terminal of claim 1, wherein the deep learning model is configured to generate the second sensing data using the feature data and data stored in a database which is determined to have a similarity with the feature. Attorney Docket No. 2060-5810

### Claim 8 (depends on 1)

The AI terminal of claim 1, wherein:
the AI function is related to image recognition for autonomous movement of the AI device, and the feature includes at least one of an illuminance of an image, a position of an object depicted in the image, a color of the object, a shape of the object, a directional position of the object, or a relative positioning between the object and a background of the image.

### Claim 9 (depends on 1)

The AI terminal of claim 1, wherein:
the AI function is related to voice recognition at the AI device; and
the feature includes at least one of an intonation of speech, a speed of speech, or a gender of a speaker associated with the voice.

### Claim 10

A computer-implemented method, the method comprising:
receiving feature data from an AI device, wherein the received feature data is generated by the AI device by obtaining sensing data and compressing the sensing data while preserving a feature of the sensing data; and
inputting the received feature data to a deep learning model to obtain second sensing data for use in a recognition model related to an AI function of the AI device.

### Claim 11 (depends on 10)

The method of claim 10, wherein the second sensing data includes a representation of the feature along with the compressed sensing data.

### Claim 12 (depends on 10)

The method of claim 10, further comprising:
using the second sensing data to train the recognition model for performing the AI function; and
transmitting the trained recognition model to the AI device.

### Claim 13 (depends on 11)

The method of claim 11, wherein the feature data is received from the AI device when the AI device fails to perform the AI function using the sensing data.

### Claim 14 (depends on 10)

The method of claim 10, wherein:
an encoder of a variational auto-encoder is used to compress the sensing data, and Attorney Docket No. 2060-5810 the deep learning model is a decoder of the variational auto-encoder.

### Claim 15 (depends on 10)

The method of claim 10, wherein the deep learning model is a generator of a generative adversarial network (GAN).

### Claim 16 (depends on 10)

The method of claim 10, wherein the deep learning model is configured to generate the second sensing data using the feature data and data stored in a database which is determined to have a similarity with the feature.

### Claim 17 (depends on 10)

The method of claim 10, wherein:
the AI function is related to image recognition for autonomous movement of the AI device, and the feature includes at least one of an illuminance of an image, a position of an object depicted in the image, a color of the object, a shape of the object, a directional position of the object, or a relative positioning between the object and a background of the image.

### Claim 18 (depends on 10)

The method of claim 10, wherein:
the AI function is related to voice recognition at the AI device, and the feature includes at least one of an intonation of speech, a speed of speech, or a gender of a speaker associated with the voice.
