原文:語音識別算法閱讀之LC-BLSTM優化版

論文: IMPROVING LATENCY CONTROLLED BLSTM ACOUSTIC MODELS FOR ONLINE SPEECH RECOGNITION 思想: BLSTM作為當前主流的序列建模算法,在語音識別領域取得了不錯的效果。但因為BLSTM的雙向LSTM結構,在序列建模時需要用到未來的時序信息,這使得算法在流式語音識別中受到制約,不滿足流式語音識別對輸出延遲的要求 而LC ...

2020-09-16 10:11 0 589 推薦指數:

查看詳情

語音識別算法閱讀之EESEN

論文:   EESEN:END-TO-END SPEECH RECOGNITION USING DEEP RNN MODELS AND WFST-BASED DECODING ...

Wed Sep 16 07:06:00 CST 2020 0 770
語音識別算法閱讀之CTC

論文:   CTC:Connectionist Temporal Classification: Labelling Unsegmented Sequence Data with Recurrent Neural Networks 思想:   語音識別中,一般包含語音 ...

Sun Sep 13 23:36:00 CST 2020 0 752
語音識別算法閱讀之DFSMN

論文: Deep-FSMN for Large Vocabulary Continuous Speech Recognition 思想:   對於大詞匯量語音識別,往往需要更深的網絡結構,但是當FSMN[1]或cFSMN[2]的結構很深時容易引發剃度消失和爆炸問題 ...

Thu Sep 17 05:51:00 CST 2020 0 863
語音識別算法閱讀之LAS

LAS:   listen, attented and spell,Google 思想:   sequence to sequence的思想,模型分為encoder和dec ...

Mon Sep 14 00:00:00 CST 2020 0 883
語音識別算法閱讀之RNN-T-2018

論文:   EXPLORING ARCHITECTURES, DATA AND UNITS FOR STREAMING END-TO-END SPEECH RECOGNITION WITH RNN- ...

Wed Sep 16 06:26:00 CST 2020 0 2315
 
粵ICP備18138465號   © 2018-2025 CODEPRJ.COM