Role separation conference shorthand system and method based on mobile terminal

A mobile terminal, conference recording technology, applied in speech analysis, speech recognition, instruments, etc., can solve problems such as large workload, separation processing lag, role separation accuracy discount, etc., to achieve fast text return and real-time correction. , transfer fast effect

Pending Publication Date: 2020-12-08
ANHUI SEMXUM INFORMATION TECH CO LTD
View PDF2 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although this method achieves voice role separation to a certain extent, it needs to configure multi-array directional microphones and various audio processing modules, and the segmentation and labeling of side audio bring a large workload to post-processing, which is not convenient Direct generation of meeting minutes text
[0004] The invention patent application CN111105801A published by the State Intellectual Property Office on May 5, 2020 discloses a method for character voice separation, which collects and organizes voice fragments based on voiceprint recognition, but as mentioned in the background technology of the invention patent CN108564952B , the separation effect of voiceprint recognition is better in an ideal recording environment, but in a more complex meeting scene, the accuracy of role separation will be greatly reduced, and post-clustering processing is required, so it is not convenient to directly generate meeting record text
[0005] The current voice role separation is mainly an independent voice separation device based on a collection of software and hardware. There is also a lag in separation processing, and it cannot be well integrated with the meeting shorthand system that requires high real-time performance to form a meeting record that can achieve role separation. system

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Role separation conference shorthand system and method based on mobile terminal
  • Role separation conference shorthand system and method based on mobile terminal

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] A mobile terminal-based meeting shorthand system for role separation, such as figure 1 As shown, it includes a mobile terminal located in front of each participant speaker, a conference shorthand server wirelessly connected to the mobile terminal, an ASR server connected to the conference shorthand server network, and an NLP server connected to the ASR server network .

[0028] The equipment for recording and preprocessing the conference audio on the mobile terminal conference site can be the tablet, mobile phone, etc. brought by the speakers, or the electronic device provided by the conference organizer. The mobile terminal ID can be the MAC address of the device, the SIM card number provided by the network operator, or other codes that can uniquely identify the device.

[0029] Both the ASR server and the NLP server are existing third-party servers. The ASR server converts the content of the audio segment into a text. This conversion process is a mechanical conversi...

Embodiment 2

[0034] Although the accuracy rate of the secondary text returned by the NLP server to the meeting shorthand server can reach 90-95%, there is still a certain error rate.

[0035] In response to this problem, this embodiment proposes a real-time manual correction scheme for the conference record text, which cuts the collected audio streams according to natural sentences through the mobile terminal, and sends the cut audio segments together with their own IDs to the The ASR server; the conference shorthand server is connected to a manual editing terminal.

[0036] The reason for cutting the audio stream is that there is a pause when people speak normally, and the natural sentence in this embodiment refers to the sentence between adjacent pauses, such as figure 2"My voice as rough as the Yellow River" and "Not only resounded in the United Nations building" in the song. Cutting the audio stream according to natural sentences can ensure the integrity of the audio information and ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a role separation conference shorthand system and method based on a mobile terminal. Voice collection is performed through the attribute of an ID of the mobile terminal, and the ID always follows the subsequent processing of an audio collected by the mobile terminal, thereby achieving audio self-tagging, and achieving role self-separation in a conference record text.

Description

technical field [0001] The invention relates to the technical field of meeting records, in particular to a mobile terminal-based role separation meeting shorthand system and method. Background technique [0002] The meeting voice is stored and converted into meeting record text in real time, gradually replacing manual meeting shorthand. However, the traditional meeting shorthand system is only used to collect speech and convert it into text, and cannot distinguish the speech of different speakers, forming a role-separated meeting record text. [0003] The invention patent CN108564952B authorized and announced by the State Intellectual Property Office on June 7, 2019 discloses a voice role separation method, which collects the voices of different people through multi-array directional microphones, and uses the combination of algorithms and hardware to improve the voice role. The accuracy of the separation, and enhance the audio of each channel audio corresponding to the spea...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G10L15/04G10L15/26G10L15/30G10L17/02G10L17/04G10L17/08G10L25/24G10L25/27G10L25/69
CPCG10L15/04G10L15/30G10L17/02G10L17/04G10L17/08G10L25/24G10L25/27G10L25/69
Inventor 虞焰兴
Owner ANHUI SEMXUM INFORMATION TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products