Error correction method and system for real-time translated text, storage medium and device

Abstract

The invention discloses an error correction method and system for a real-time translated text, a storage medium and a device, and relates to the field of voice error correction. The method comprises the steps that: an ASR translation text in real-time live broadcast is acquired, the ASR translation text is interpreted through a trained BERT error correction model, and a first error correction text is output; the ASR translation text is interpreted through the trained GPT error correction model, a second error correction text is output, the first error correction text and the second error correction text are combined to obtain an error correction target text, and end-to-end error correction is performed on the ASR translation text content in the live broadcast scene through the scheme, the character accuracy of the ASR on an audio translated text can be effectively improved, and the method can be quickly applied to the field of live broadcast.

Description

technical field [0001] The invention relates to the field of speech error correction, in particular to an error correction method, system, storage medium and device for real-time translated text. Background technique [0002] Due to the large storage capacity and complex content of voice information, it is not easy to store, monitor and analyze directly. Therefore, automatic speech recognition technology (ASR) is used for voice-text translation, and the text is further stored, monitored and analyzed. [0003] In recent years, with the rise of the webcast industry, a large amount of information with voice signals as the carrier has been disseminated on the Internet. When using ASR for speech-to-text translation, due to the uneven quality of the live broadcast environment and the insufficient capacity of the ASR model, it is enough to change the semantics. Mistranslated information, such as translating the audio information of "I want to go to Dali" into the text information o...

AI Technical Summary

Problems solved by technology

On the one hand, the defect of this type of algorithm is that when the error correction position of the detector mark is wrong, additional errors will be added.

On the other hand, the maintenance of the two-stage error correction method is cumbersome, especially the construction of the candidate set of the error corrector. The spoken language interaction in the live broadcast scene is complex, there are many corpus, and it is more difficult to explicitly count ASR translation errors.

At present, in the live broadcast scene, the ASR error correction method of the traditional language model is difficult to reach the use level

Images