Voice Activity Detection for Communication Session Transcription
Solution Overview
Problem
Telephone systems may fail to register the end of communication sessions due to the absence of voltage swings or other indicators, leading to operational problems.
Innovation Solution
A method that transcribes communications by obtaining audio during a session, establishing a network connection with a transcription system, and monitoring for inactive voice to stop sending audio or terminate the connection, thereby reducing network traffic and processing requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If audio is continuously sent to the transcription system during a communication session, then complete transcription coverage is achieved, but network traffic and processing requirements increase
Solution Approach 1:
The system uses voice activity detection to periodically assess whether voice is present in the audio stream, and only sends audio to the transcription system when voice activity is detected. This periodic monitoring and conditional transmission reduces network traffic while maintaining reliable transcription coverage for actual speech content.
Solution Approach 2:
The system extracts and identifies voice activity from the continuous audio stream using voice activity detection. By separating voice-containing segments from non-voice segments, the system transmits only the necessary audio portions to the transcription system, reducing overall network traffic while preserving complete transcription coverage for spoken content.
2Ease of operation
If the communication session is terminated when voltage swing is not detected, then session management is simplified, but sessions may be incorrectly ended
Solution Approach 1:
The system uses voice activity detection as a feedback mechanism to continuously monitor the communication session status. Instead of relying solely on voltage swing detection, the system actively monitors for voice activity and uses this feedback to determine when to terminate the session, reducing incorrect session endings while maintaining simplified session management.
Solution Approach 2:
The system replaces the traditional voltage swing detection mechanism with voice activity detection. This substitution uses signal processing and pattern recognition to identify voice activity more reliably, preventing incorrect session terminations while maintaining ease of session management through automated detection.
3Productivity
If voice activity monitoring is implemented to stop audio sending, then network efficiency improves, but system complexity increases
Solution Approach 1:
The system implements self-service through automated voice activity detection that automatically determines when to start and stop audio transmission to the transcription system. This self-managed monitoring and control mechanism improves network efficiency by eliminating unnecessary transmissions while adding minimal complexity through integrated detection algorithms.
Data Source
AI summary
A method to transcribe communications may include obtaining, during a communication session, audio that includes a voice of a user participating in the communication session. The communication session may be configured for verbal communication. The method may further include establishing a network connection with a transcription system and sending the audio to the transcription system. In some embodiments, the transcription system may be configured to generate a transcript of the audio. The method may also include obtaining the transcript of the audio from the transcription system during the communication session and monitoring the audio to determine when the voice is inactive. In some embodiments, in response to the voice being inactive, the method may include stopping the sending of the audio to the transcription system while maintaining the communication session.