Synchronized Caption Generation Through Transcript Timing Correlation

Solution Overview

Problem

Existing captioning technologies struggle with synchronization accuracy, whether relying on human transcribers or automated methods, leading to captions being out of sync with audio.

Innovation Solution

Correlate timing information from an automatically generated first transcript with a second transcript, potentially created by human transcribers, to synchronize captions accurately by comparing word matches and using multiple transcriptions for improved accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If automated transcription techniques are used, then timing information accuracy is improved, but transcription accuracy deteriorates

Engineering Contradiction:

Improvetiming information accuracyVSAvoidtranscription accuracy

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The patent combines automated transcription techniques (which provide accurate timing information) with human transcriber input (which provides accurate transcription). The system receives both automated transcripts with timing data and human-transcribed text, then merges these two data sources to produce final captions that have both accurate timing and accurate transcription content.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent uses an intermediary processing system that takes automated transcription output and human transcriber input as separate inputs, processes them together, and produces the final synchronized captions. This intermediary system correlates the timing information from automated transcripts with the transcription accuracy from human input.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If human transcribers are used, then transcription accuracy is improved, but timing information accuracy deteriorates

Engineering Contradiction:

Improvetranscription accuracyVSAvoidtiming information accuracy

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The patent combines automated transcription techniques (which provide accurate timing information) with human transcriber input (which provides accurate transcription). The system receives both automated transcripts with timing data and human-transcribed text, then merges these two data sources to produce final captions that have both accurate timing and accurate transcription content.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent uses an intermediary processing system that takes automated transcription output and human transcriber input as separate inputs, processes them together, and produces the final synchronized captions. This intermediary system correlates the timing information from automated transcripts with the transcription accuracy from human input.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If multiple transcribers are used, then transcription accuracy is improved, but device complexity increases

Engineering Contradiction:

Improvetranscription accuracyVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent uses multiple transcribers (human and automated) to create multiple versions of the transcript. These copies are then processed through a correlation system that compares and synthesizes them into a final accurate transcript. The multiple copies allow the system to leverage diverse transcription styles and accuracy levels.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent combines multiple transcription sources (human and automated) into a unified final transcript. The system merges the strengths of different transcribers by correlating their outputs and producing a single accurate result that benefits from the diversity of multiple perspectives.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS12451154B2Generating and/or displaying synchronized captions

Publication Date: 2025.10.21 COMCAST CABLE COMM LLC

AI summary

Methods, apparatuses, and systems are described for correlating automatically-generated timing information from a first audio transcript with a second audio transcript (such as a transcript generated by one or more human transcribers) that may not have timing information. By correlating the second transcript with the timing information, an accurate and synchronized transcript may be generated. To correlate the second transcript with the timing information, a first transcript that contains the timing information may be automatically generated, and words of the first transcript may be compared to words of the second transcript. Based on the comparison, the timing information of the first transcript can be associated with the words of the second transcript.