Sensor-Based Play Event Detection in Self-Recorded Sports

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing technologies face difficulties in automatically detecting desired scenes in sports content recorded by users themselves, as they lack features like whistles or shouts, making it harder to learn and detect play events compared to commercial content.

Innovation Solution

An information processing device and system that uses sensor devices to acquire user behavior data and generate play event information, allowing for the detection and recording of predefined play events in sports, such as movements or swings, to enhance the analysis and review of personal sports performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If feature amounts of images and sounds are extracted and learned to detect highlight scenes, then commercial content analysis is effective, but self-recorded sports content cannot be properly analyzed due to lack of characteristic features

Engineering Contradiction:

Improvescene detection accuracyVSAvoidapplicability to self-recorded content

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent changes the detection parameters from commercial content features (whistles, shouts, field lines) to self-recorded content features (user behavior patterns, movement trajectories, acceleration profiles). This allows the same detection system to work effectively across different content types by adapting which parameters are monitored and how they are interpreted.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent introduces sensor information as an intermediary that bridges the gap between self-recorded content and meaningful play event detection. Sensors mounted on the user or equipment provide objective behavioral data that serves as a reliable mediator for detecting play events without requiring traditional commercial content features.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If traditional scene detection methods are used on self-recorded content, then some detection may occur, but detection accuracy deteriorates due to unnecessary scenes and lack of characteristic features

Engineering Contradiction:

Improvereview efficiencyVSAvoidplay event detection accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments the continuous video content into discrete play events based on detected user behaviors. By dividing the content into meaningful units (serves, rallies, faults) identified through sensor-triggered detection, the system improves both review efficiency and detection accuracy, allowing users to quickly navigate to specific play events rather than reviewing entire matches.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system uses sensor information as feedback to continuously monitor and detect play events in real-time. The sensor data provides immediate feedback about user behavior that triggers automated detection and marking of play events, enabling accurate identification without manual review and significantly improving both productivity and precision.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11250247B2Information processing device, information processing system, and program

Publication Date: 2022.02.15 SONY GROUP CORP

AI summary

There is provided an information processing device including a control unit to generate play event information based on a determination whether detected behavior of a user is a predetermined play event.