# US 2026/0289268 A1 · System

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

- Number: US 2026/0289268 A1
- Document ID: US20260289268A1
- Status: Published application
- Kind: A1
- Application number: 19/565,640
- Filed: 2026-03-13
- Published: 2026-09-24
- Earliest claimed: 2025-03-19
- Assignees: SoftBank Group Corp.
- Applicants: SoftBank Group Corp.
- Inventors: Natsumi KOBAYASHI
- Web page: https://patentgrep.com/patent/US20260289268A1

## CPC

- G06N 3/0475: Generative networks
- G06F 40/242: Dictionaries
- G06F 40/247: Thesauruses; Synonyms
- G06F 40/253: Grammatical analysis; Style critique
- G06F 40/284: Lexical analysis, e.g. tokenisation or collocates
- G06F 40/295: Named entity recognition
- G06F 40/40: Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30)
- G06V 10/7715: {Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods}
- G06V 10/82: using neural networks
- G06V 20/70: Labelling scene content, e.g. deriving syntactic or semantic representations
- G10L 13/027: Concept to speech synthesisers; Generation of natural phrases from machine-based concepts (generation of parameters for speech synthesis out of text G10L13/08)
- 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

A system includes a processor that is configured to analyze prompt data input by a user for instructing a generative AI model to execute a specific processing operation, and search for visual information related to the prompt data by using the generative AI model, analyze the visual information by using the generative AI model and generate a descriptive sentence that explains content of the visual information, and generate, by using the generative AI model and based on the generated descriptive sentence, a summary sentence including information that an advertiser desires to add.

## Claims

### Claim 1

A system comprising:
circuitry configured to:
receive, via a communication interface coupled to a packet-switched network, prompt data from a terminal device operated by a user, and analyze the prompt data using a natural language processing resource to extract feature terms;
retrieve, based on the extracted feature terms, visual information from an information source via the packet-switched network, and select visual information from a plurality of pieces of acquired visual information in accordance with a relevance criterion;
input the selected visual information to an image analysis resource and acquire analysis result data comprising object information and scene information;
construct a prompt data structure for a generative neural network model based on the analysis result data, and cause the generative neural network model to generate a descriptive output sentence describing content of the visual information;
cause the generative neural network model to generate, based on the descriptive output sentence and supplemental attribute data stored in a storage device, a composite output sentence incorporating the supplemental attribute data; and
transmit, via the communication interface, the composite output sentence and the visual information to the terminal device for presentation via a user interface of the terminal device.

### Claim 2 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to estimate an emotional state of the user based on at least one of expression data, audio signal data, or interaction pattern data received from the terminal device, and to adjust at least one of a content, a style, or a presentation order of the composite output sentence in accordance with the estimated emotional state.

### Claim 3 (depends on 2)

The system according to claim 2, wherein the circuitry is configured to estimate the emotional state using an emotion recognition resource comprising a neural network classifier that outputs a probability distribution over a plurality of discrete emotional state categories.

### Claim 4 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to store the descriptive output sentence, the composite output sentence, and the visual information in association with identification information in the storage device, and to transmit the stored data to a content distribution apparatus via the packet-switched network.

### Claim 5 (depends on 1)

The system according to claim 1, wherein the natural language processing resource performs tokenization, part-of-speech tagging, and named entity recognition on the prompt data, and wherein the extracted feature terms comprise nouns and named entities with stop words removed.

### Claim 6 (depends on 5)

The system according to claim 5, wherein the circuitry is configured to normalize the extracted feature terms by applying lemmatization, case conversion, and synonym mapping using a predefined dictionary prior to retrieving the visual information.

### Claim 7 (depends on 1)

The system according to claim 1, wherein the image analysis resource comprises a convolutional neural network or a transformer-based vision model that generates feature maps from pixel data of the selected visual information and outputs object labels, scene labels, and confidence scores.

### Claim 8 (depends on 7)

The system according to claim 7, wherein the circuitry is configured to filter the object labels and scene labels by retaining labels having confidence scores exceeding a predetermined threshold, and to group retained labels into categories to form the analysis result data.

### Claim 9 (depends on 1)

The system according to claim 1, wherein constructing the prompt data structure comprises converting the object information and scene information into textual fragments, combining the textual fragments with instruction parameters specifying a language, a style, and a length constraint for the descriptive output sentence.

### Claim 10 (depends on 9)

The system according to claim 9, wherein the generative neural network model comprises a transformer architecture with self-attention layers, and wherein the circuitry is configured to control inference parameters comprising at least one of a temperature parameter, a top-k sampling threshold, or a maximum output length during generation of the descriptive output sentence.

### Claim 11 (depends on 1)

The system according to claim 1, wherein causing the generative neural network model to generate the composite output sentence comprises constructing a second prompt data structure including the descriptive output sentence as contextual information and the supplemental attribute data as content to be incorporated, and transmitting the second prompt data structure to the generative neural network model.

### Claim 12 (depends on 11)

The system according to claim 11, wherein the second prompt data structure includes instructions specifying that the generative neural network model should output a concise sentence that incorporates the supplemental attribute data in a contextually coherent manner with respect to the descriptive output sentence.

### Claim 13 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to perform post-processing on the composite output sentence comprising validation against a maximum character count, detection and removal of prohibited terms, and style normalization.

### Claim 14 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to perform post-processing on the descriptive output sentence comprising trimming whitespace, ensuring terminal punctuation, and verifying that a length of the descriptive output sentence does not exceed a predetermined limit.

### Claim 15 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to set the composite output sentence as alternative text in attribute information of a structured document associated with the visual information.

### Claim 16 (depends on 15)

The system according to claim 15, wherein the circuitry is configured to input the composite output sentence to a speech synthesis resource to generate an audio signal, and to transmit the audio signal to the terminal device for auditory output.

### Claim 17 (depends on 1)

The system according to claim 1, wherein the circuitry is configured to store history information comprising the prompt data, the analysis result data, the descriptive output sentence, and the composite output sentence in the storage device, and to update at least one of a model selection condition and a generation condition parameter for a subsequent generation process based on the stored history information.

### Claim 18

A system comprising:
circuitry configured to:
receive, via a communication interface coupled to a packet-switched network, prompt data from a terminal device, and analyze the prompt data using a natural language processing resource to extract feature terms;
retrieve visual information based on the extracted feature terms from an information source via the packet-switched network, input the visual information to an image analysis resource comprising a convolutional neural network or a transformer-based vision model, and acquire analysis result data comprising object labels, scene labels, and confidence scores;
construct a first prompt data structure based on the analysis result data and cause a generative neural network model comprising a transformer architecture with self-attention layers to generate a descriptive output sentence describing content of the visual information;
construct a second prompt data structure comprising the descriptive output sentence and supplemental attribute data stored in a storage device, and cause the generative neural network model to generate a composite output sentence incorporating the supplemental attribute data;
estimate an emotional state of a user of the terminal device and adjust at least one of a content or a presentation order of the composite output sentence based on the estimated emotional state; and
transmit the composite output sentence and the visual information to the terminal device for presentation via a user interface of the terminal device.

### Claim 19 (depends on 18)

The system according to claim 18, wherein the circuitry is configured to set the composite output sentence as alternative text in attribute information of a structured document associated with the visual information, and to input the composite output sentence to a speech synthesis resource to generate an audio signal for auditory output.

### Claim 20

A method comprising:
receiving, via a communication interface coupled to a packet-switched network, prompt data from a terminal device operated by a user, and analyzing the prompt data using a natural language processing resource to extract feature terms;
retrieving, based on the extracted feature terms, visual information from an information source via the packet-switched network, and selecting visual information from a plurality of pieces of acquired visual information in accordance with a relevance criterion;
inputting the selected visual information to an image analysis resource and acquiring analysis result data comprising object information and scene information;
constructing a prompt data structure for a generative neural network model based on the analysis result data, and causing the generative neural network model to generate a descriptive output sentence describing content of the visual information;
causing the generative neural network model to generate, based on the descriptive output sentence and supplemental attribute data stored in a storage device, a composite output sentence incorporating the supplemental attribute data; and
transmitting, via the communication interface, the composite output sentence and the visual information to the terminal device for presentation via a user interface of the terminal device.
