Network teaching method and system with voice assessment function

Abstract

The invention provides a network teaching method and system with a voice assessment function. According to the voice assessment method provided by the invention, a phoneme state of a voice is used for replacing a multi-Gaussian mixture model trained by a conventional Mel-frequency cepstral coefficient (MFCC), and a posterior probability and a zero-order Baum-Welch statistical magnitude are calculated according to the feature. A voice feature based on phonemes is extracted through a multi-language phoneme identifier. A feature based on multi-language extraction is complementary during catching of non-native pronunciation information, and a feature based on phoneme duration is effective in automatic native accent assessment. Finally, a fusion system is provided in the method, so that Spearman relevant coefficients of 0.5706 and 0.6089 are reached on a development set and a test set. As indicated by the relevant coefficients, the method provided by the invention is very accurate and effective in oral speech assessment.

Description

technical field [0001] The present invention relates to an automatic local accent assessment method based on multilingual phoneme posterior probability and multilingual phoneme duration correlation features, which belongs to the field of speech recognition, especially relates to the field of spoken language assessment, and also relates to a network teaching with speech assessment function methods and systems. Background technique [0002] In recent years, with the popularization of the mobile Internet and the rapid development of the network teaching system, the traditional teaching mode has gradually been unable to meet the needs of users for multimedia information teaching methods. In the online teaching process, the real-time synchronous display of the teaching process, the real-time recording of the teaching process, and the fast and effective on-demand playback are more and more concerned by the majority of users. In the prior art, in order to meet the above needs, man...

AI Technical Summary

Problems solved by technology

This is because most of the existing network teaching systems use video streams and manual screenshots by teachers according to their own judgments in the process of transmitting data on the same screen. Concurrent network transmission often has slow speed, unsmooth display, and data freezes, which affect the students' experience in class, and teachers have to consider timely screenshots while lecturing, and sometimes forget to take screenshots and send them to students. Students can't judge whether they have obtained timely screenshots, which affects the teaching effect

[0012] Specifically, the shortcomings include: 1. The recording process is cumbersome, and some require special personnel to record on-site; 2. The video file is huge, and the quality of the courseware needs to be sacrificed for multiple format conversions to reduce the file size; 3. The server storage system requires high requirements; 4. Local recording requires high configuration and performance of the client machine, slow upload speed, and difficult playback; 5. The role of the simulation system consumes high server resources, which cannot meet the large concurrent network recording requirements of the classroom; 6. The performance of streaming media server clusters for on-demand playback High requirements; 7. On-demand playback requires high server bandwidth; 8. On-demand playback requires users to have high bandwidth (the file is too large); 9. When using 2G / 3G / 4G wireless networks to watch, serious freezes, not smooth enough, and costly high traffic

English-only phoneme recognizers may not be able to capture specific information from the speaker's native language

Images