Live Video Tagging with Delayed Stream Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In live video tagging, continuous scenes often fail to be tagged correctly, leading to inefficiencies as the video cannot be rewound to re-tag missed scenes, requiring time to re-produce and re-search the video for retagging.

Innovation Solution

An information processing apparatus and method that displays a live video alongside a delayed version of the same video, allowing users to input tags in real-time and correct or re-input tags on the delayed video, facilitating easy tagging by synchronizing tag inputs with the correct timing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If tagging is performed on live video in real-time, then tagging speed is improved, but tagging accuracy deteriorates due to continuous scenes causing tag failures

Engineering Contradiction:
Improvetagging speedVSAvoidtagging accuracy
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary tagging on the delayed video after the live video has been processed. This allows the tagging operation to be executed with sufficient time for review and correction, ensuring accuracy while maintaining real-time processing capability for the live stream.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

A delayed video stream is introduced as an intermediary between the live video input and the final tagged output. This intermediary allows the tagging system to review and correct tags before they are finalized, resolving the contradiction between speed and accuracy by providing a buffer for quality control.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If tagging fails on continuous scenes, then tagging reliability deteriorates, but time loss increases due to inability to rewind and re-tag

Engineering Contradiction:
Improvetagging reliabilityVSAvoidtime for retagging
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The delayed video provides a time buffer that allows failed tags to be corrected in the delay period before the video is finalized. This preliminary correction opportunity eliminates the need for time-consuming rewinding and re-searching, as all corrections can be made while the video is still in the delayed state.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system builds in a time cushion through the delayed video stream, creating a safety margin that accommodates tagging failures. This beforehand cushioning ensures that even if tagging fails on continuous scenes, there is sufficient time to correct errors without requiring video rewinding or causing production delays.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

3Measurement precision

If a delayed video is displayed alongside live video, then tagging accuracy is improved through correction opportunities, but device complexity increases

Engineering Contradiction:
Improvetagging precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The delayed video stream serves multiple functions: it acts as a review buffer for tag correction, provides a backup for failed tags, and enables quality control without requiring separate complex systems. This multi-functionality reduces overall system complexity while maintaining high tagging precision.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

Instead of creating entirely separate tagging systems for live and delayed videos, the system uses a simplified copy of the video stream with the same tagging interface. This copying approach allows tag correction on the delayed stream using the same tools as the live stream, reducing complexity while improving precision.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS11694725B2Information processing apparatus and information processing method
Publication Date: 2023.07.04 SONY GROUP CORP
  • US11694725B2 patent drawing
  • US11694725B2 patent drawing
  • US11694725B2 patent drawing

AI summary

This information processing apparatus includes displaying video content on a first display region in a display section as a first video, displaying the above-mentioned video content on a second display region in the above-mentioned display section as a second video delayed from the above-mentioned first video by a predetermined time, and setting a first tag inputted by a user into the above-mentioned first video and a second tag inputted by the above-mentioned user into the above-mentioned second video as tags for the above-mentioned video content.