Multimedia Device Speech Recognition Context Integration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current speech recognition systems primarily focus on recognizing spoken content rather than determining the user's intention, and they do not effectively utilize contextual information such as time and ambient environment to provide tailored services, especially in multimedia devices like smart TVs and mobile phones.

Innovation Solution

The system enhances speech recognition by incorporating contextual information, including the time and ambient environment, to differentiate user intentions and provide personalized services, using a multimedia device that captures video data and transmits it to servers for natural language processing and image recognition to accurately interpret user commands.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If speech recognition systems focus only on recognizing spoken content, then recognition accuracy is improved, but the ability to detect user intention and provide personalized services deteriorates

Engineering Contradiction:
Improvespeech recognition accuracyVSAvoiduser intention detection capability
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent combines speech recognition with multiple contextual information sources including time information, ambient environment data, and application execution status. The controller integrates these diverse data types to comprehensively determine user intention, merging the strengths of speech recognition with environmental awareness and application context to resolve the contradiction between recognition accuracy and intention detection capability.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent adds new dimensions to speech recognition by incorporating temporal dimension (time information), spatial dimension (ambient environment), and contextual dimension (application execution status). This multi-dimensional approach transforms the system from simple speech-to-text conversion to comprehensive intention understanding, addressing the limitation of traditional speech recognition systems.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If contextual information is incorporated to detect user intention, then service personalization is improved, but system complexity increases

Engineering Contradiction:
Improveservice personalization capabilityVSAvoidsystem structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the intention detection process into distinct functional modules: a time information acquisition module, an ambient environment detection module, an application status monitoring module, and a controller that integrates these inputs. This segmentation allows each module to handle specific aspects of contextual information independently, making the overall complex system more manageable and maintainable while achieving comprehensive intention detection.

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If multiple data sources are integrated for intention detection, then detection precision is improved, but processing time increases

Engineering Contradiction:
Improveintention detection precisionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent implements preliminary action by continuously acquiring and pre-processing contextual information (time, ambient environment, application status) in the background before speech recognition is triggered. This allows the system to have contextual data readily available when speech input occurs, reducing the processing time required for intention detection while maintaining high precision through the integration of multiple pre-acquired data sources.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3660841B1Multimedia device for processing voice command
Publication Date: 2023.08.16 LG ELECTRONICS INC
  • EP3660841B1 patent drawingFigure 1
  • EP3660841B1 patent drawingFigure 2
  • EP3660841B1 patent drawingFigure 3

AI summary

The present invention discloses a multimedia device capable of processing a speech-based command. One embodiment of the present invention provides a multimedia device including a memory to store at least one application therein; an application manager for executing any application among the at least one application stored in the memory; and a controller configured to receive a speech-based data from an outside, wherein the controller is configured: to capture video data from a currently-executed application in response to the received speech-based data; to control a network interface module to transmit to a server the captured video data, the received speech-based data, and additional information about the currently-executed application; and to control the network interface module to receive a feedback result value associated with the speech-based data from the server, wherein the feedback result value varies for the same speech-based data based on the captured video data and the additional information about the currently-executed application.