# US 12,217,485 B2 · Object recognition device, method, and computer-readable medium

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

- Number: US 12,217,485 B2
- Document ID: US12217485B2
- Status: Granted
- Kind: B2
- Application number: 17/768,597
- Filed: 2019-10-24
- Published: 2024-04-04
- Granted: 2025-02-04
- Earliest claimed: 2019-10-24
- Assignees: NEC Corporation
- Applicants: NEC Corporation
- Inventors: Azusa Sawada; Soma Shiraishi; Takashi Shibata
- Web page: https://patentgrep.com/patent/US12217485B2

## CPC

- G06V 10/7715: {Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods}
- G06F 16/55: Clustering; Classification
- G06F 18/24147: {Distances to closest patterns, e.g. nearest neighbour classification}
- G06T 7/00: Image analysis
- G06V 10/772: Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
- G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
- G06V 10/778: Active pattern-learning, e.g. online learning of image or video features
- G06V 40/10: Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V 10/82: using neural networks
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

The learning device includes a metric space learning unit and a case example storage unit. The metric space learning unit learns a metric space including feature vectors extracted from attributed image data, for each combination of different attributes, using the attributed image data to which attribute information is assigned. The case example storage unit computes the feature vector from the image data for case example to store the computed feature vector as a case example associated with the metric space, and stores additional information associated with the case example.

## Claims

### Claim 1

An inference device comprising:
a memory configured to store instructions; and
one or more processors configured to execute the instructions to:
store a feature vector of image data for case example in association with a metric space learned for each combination of different attributes, and store additional information associated with the case example by a learning method;
wherein the learning method comprises:
learning a metric space including feature vectors extracted from attributed image data, for each combination of different attributes, using the attributed image data to which attribute information is assigned; and
computing the feature vector from the image data for case example to store the computed feature vector as a case example associated with the metric space, and storing additional information associated with the case example;
identify image data for inference based on the feature vector extracted from the image data for inference and the case example associated with the metric space; and
generate, using the additional information associated with the case example used for identification, an identification result.

### Claim 2 (depends on 1)

The inference device according to claim 1, the one or more processors are further configured to acquire a plurality of metric spaces from the case example storage and evaluate the plurality of metric spaces using the feature vectors of the image data for selection to select one metric space,
wherein the one or more processors identify the image data for inference based on the feature vector extracted from the image data for inference and the case example associated with the one metric space.

### Claim 3 (depends on 1)

The inference device according to claim 1,
wherein the additional information includes representative image data corresponding to the case example, and
wherein the one or more processors are further configured to execute the instructions to display the identification result and the representative image data on a display device.

### Claim 4 (depends on 1)

The inference device according to claim 1, wherein the additional information includes any one of a label, an imaging condition, and environmental information of the image data for case example.

### Claim 5 (depends on 1)

The inference device according to claim 1, wherein the additional information includes representative image data corresponding to the case example.

### Claim 6 (depends on 5)

The inference device according to claim 5, wherein the additional information includes any one of a label, an imaging condition, and environmental information of the image data for case example.

### Claim 7

An inference method comprising:
acquiring a metric space from a case example storage which stores a feature vector of image data for case example in association with a metric space learned for each combination of different attributes, and which stores additional information associated with the case example, the case example storage being created by a learning method;
wherein the learning method comprises:
learning a metric space including feature vectors extracted from attributed image data, for each combination of different attributes, using the attributed image data to which attribute information is assigned; and
computing the feature vector from the image data for case example to store the computed feature vector as a case example associated with the metric space, and storing additional information associated with the case example;
identifying image data for inference based on a feature vector extracted from the image data for inference and the case example associated with the metric space; and
generating, using the additional information associated with the case example used for identification, an identification result.

### Claim 8 (depends on 7)

The inference method according to claim 7, wherein the additional information includes representative image data corresponding to the case example.

### Claim 9 (depends on 8)

The inference method according to claim 8, wherein the additional information includes any one of a label, an imaging condition, and environmental information of the image data for case example.

### Claim 10

A non-transitory computer-readable recording medium recording a program causing a computer to execute processing of:
acquiring a metric space from a case example storage unit-which stores a feature vector of image data for case example in association with a metric space learned for each combination of different attributes, and which stores additional information associated with the case example, the case example storage being created by a learning method;
wherein the learning method comprises:
learning a metric space including feature vectors extracted from attributed image data, for each combination of different attributes, using the attributed image data to which attribute information is assigned; and
computing the feature vector from the image data for case example to store the computed feature vector as a case example associated with the metric space, and storing additional information associated with the case example;
identifying image data for inference based on a feature vector extracted from the image data for inference and the case example associated with the metric space; and
generating, using the additional information associated with the case example used for identification, an identification result.

### Claim 11 (depends on 10)

The recording medium according to claim 10, wherein the additional information includes representative image data corresponding to the case example.

### Claim 12 (depends on 11)

The recording medium according to claim 11, wherein the additional information includes any one of a label, an imaging condition, and environmental information of the image data for case example.