Recording processing method and device

Abstract

Embodiments of the present invention provide a recording processing method and device, which belong to the technical field of speech processing. Including: for any sound card among the N sound cards whose calibrated data is aligned, if channel t is used as the main channel in the preset number of channels of any sound card, then obtain the duration of channel t as the main channel; according to the duration , determine a target channel from the preset number of channels of any sound card, and use the voice signal collected by the target channel within the duration period as the only voice signal recorded by any sound card within the duration period, the duration period is the time period when channel t is used as the main channel. Because it can avoid the problem that the content of the speech is confused because a certain speaker speaks to a certain channel, and other channels also collect the speech of the speaker. At the same time, one-to-one correspondence between the recorded speech content and the speaker can be achieved, so as to achieve role separation, so as to facilitate subsequent review and listening efficiently.

Description

technical field [0001] The invention relates to the technical field of speech processing, in particular to a recording processing method and device. Background technique [0002] At present, the judiciary needs to record the trial situation and the content of the question and answer during the trial process, which requires considerable manpower. The development of speech recognition technology provides an effective way to solve this problem. In the related art, recording is generally performed directly through common recording equipment. However, the acoustic phenomenon in the actual trial environment is more complicated: First, for some trial scenes with large spaces that cause severe reverberation, environmental noise such as air conditioners, power amplifiers, and auditor noise will make the recording signal-to-noise ratio low. Secondly, the speakers generally include different roles such as judge, defender, plaintiff and defendant, and the distance between them is relat...

AI Technical Summary

Problems solved by technology

However, the acoustic phenomenon in the actual trial environment is more complicated: First, for some trial scenes where the space is large and the reverberation is serious, environmental noise such as air conditioners, power amplifiers and audition noise will make the recording signal-to-noise ratio low

Secondly, the speakers generally include different roles such as judge, defender, plaintiff and defendant, and the distance between them is relatively scattered. The recording equipment cannot take care of everyone, and there will be large reverberation in the recording data, which greatly reduces the clarity and intelligibility of the recording. Spend

Finally, there are many people speaking in the trial, and the contents of the speeches recorded by ordinary recording equipment are mixed together, which makes it time-consuming to review and listen to the trial records and recordings.

Images