淺析AI語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯上應(yīng)用的可能性_第1頁(yè)
淺析AI語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯上應(yīng)用的可能性_第2頁(yè)
淺析AI語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯上應(yīng)用的可能性_第3頁(yè)
淺析AI語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯上應(yīng)用的可能性_第4頁(yè)
淺析AI語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯上應(yīng)用的可能性_第5頁(yè)
免費(fèi)預(yù)覽已結(jié)束,剩余2頁(yè)可下載查看

下載本文檔

版權(quán)說(shuō)明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡(jiǎn)介

1、    淺析ai語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯上應(yīng)用的可能性    楊茜 袁奧航 胡歡 袁玉 劉鈞鵬 朱奕 阮先玉摘要:隨著全球化的發(fā)展,我國(guó)與國(guó)外文化交流日益頻繁,英文視頻的需求量大幅上升。ai語(yǔ)音識(shí)別技術(shù)的應(yīng)用極大的促進(jìn)了語(yǔ)言產(chǎn)業(yè)的創(chuàng)新。為研究ai語(yǔ)音識(shí)別技術(shù)在傳統(tǒng)聽譯工作中應(yīng)用的可能,本文同時(shí)使用訊飛聽見、騰訊云、搜狗聽寫三個(gè)支持語(yǔ)音識(shí)別技術(shù)的軟件,對(duì)人工聽譯與ai語(yǔ)音識(shí)別聽譯后的文本進(jìn)行了初步分析與總結(jié)。本文發(fā)現(xiàn),ai語(yǔ)音識(shí)別較人工聽譯用時(shí)短,但正確率有待提高,就如何對(duì)兩者的優(yōu)缺點(diǎn)進(jìn)行結(jié)合,本文提出了相應(yīng)思路和方法。關(guān)鍵詞:聽譯;ai語(yǔ)音識(shí)別;語(yǔ)音轉(zhuǎn)寫在

2、“引進(jìn)來(lái)”和“走出去”戰(zhàn)略的指導(dǎo)下,我們對(duì)英文視頻的需求量日益增加。聽譯是指對(duì)音頻或視頻中的原聲語(yǔ)音文本進(jìn)行聽寫和識(shí)別,便于后續(xù)對(duì)音頻或視頻進(jìn)行翻譯的過(guò)程。傳統(tǒng)人工聽譯依靠人工提取,對(duì)速記員要求較高,受人為因素影響較大。隨著人工智能技術(shù)的日漸成熟,ai語(yǔ)音識(shí)別技術(shù)在語(yǔ)音識(shí)別和聽寫方面受到更廣泛的認(rèn)可。2017年8月,微軟宣布其旗下語(yǔ)音識(shí)別系統(tǒng)的正確率已經(jīng)由原來(lái)的94.1%提升至94.9%,其正確率高于部分專業(yè)速記員。然而在語(yǔ)音特征提取的準(zhǔn)確性,識(shí)別的穩(wěn)定性等方面亟待改進(jìn)。1.傳統(tǒng)人工聽譯的特點(diǎn)及問(wèn)題聽譯是一種特殊的語(yǔ)音識(shí)別和轉(zhuǎn)換類型,具有書面性,即時(shí)性,同步性,跨文化性等特性。針對(duì)英文視頻的語(yǔ)

3、音識(shí)別,聽譯時(shí)并無(wú)源語(yǔ)文本作為參考。完成從音頻到書面文本的轉(zhuǎn)換,要求速記員有較高的聽辨能力。然而,英文音頻源文本具有口語(yǔ)化、不規(guī)范性、難以識(shí)別性等特征,使得速記員在聽譯時(shí)很難辨識(shí)。2.ai語(yǔ)音識(shí)別聽譯與人工聽譯的分析與比較選用音視頻均來(lái)自ted演講、bbc新聞、知名電影片段,ai語(yǔ)音識(shí)別軟件采用訊飛聽見、騰訊云、搜狗聽寫三個(gè)支持ai語(yǔ)音識(shí)別(語(yǔ)音轉(zhuǎn)文字)的軟件。2.1用時(shí)以ted演講如何學(xué)好外語(yǔ)為例,速記員人工聽譯平均用時(shí)一小時(shí)三十七分二十七秒(1:37:27),三個(gè)ai語(yǔ)音識(shí)別軟件平均用時(shí)十一分零九秒(11:09),ai軟件語(yǔ)音識(shí)別并生成文本幾乎與原視頻同步。對(duì)比之下,筆者組織速記員對(duì)50個(gè)

4、不同音頻進(jìn)行人工聽譯,并對(duì)用時(shí)進(jìn)行統(tǒng)計(jì)。統(tǒng)計(jì)結(jié)果顯示,人工聽譯文本的用時(shí)是ai語(yǔ)音識(shí)別軟件的3-14倍,倍數(shù)與源語(yǔ)文文本的時(shí)長(zhǎng)和難度呈正相關(guān)。統(tǒng)計(jì)結(jié)果表明,在用時(shí)方面,ai語(yǔ)音識(shí)別軟件體現(xiàn)出其明顯優(yōu)勢(shì)。2.2口音校正速記員在人工聽譯時(shí)能針對(duì)口音較重的音頻進(jìn)行反復(fù)多次的聽寫,從而達(dá)到終版聽譯文本的準(zhǔn)確。然而,由于大部分語(yǔ)音識(shí)別軟件默認(rèn)標(biāo)準(zhǔn)的美式或英式發(fā)音,對(duì)部分帶有口音的音頻存在識(shí)別障礙。例1:人工:.talking about how this problem is being addressed.搜狗/騰訊:.talking about how this problem is being d

5、angerous.例2:人工:. after the third season, seriously, the dialogue started to make sense.搜狗:. after they turn a season, seriously, the dialogue started to make sense.以上材料均選用帶有印式英語(yǔ)的音頻。不難發(fā)現(xiàn),由于印式英語(yǔ)與美式英語(yǔ)和英式英語(yǔ)之間存在元音障礙和輔音障礙,ai語(yǔ)音識(shí)別軟件難以對(duì)部分發(fā)音進(jìn)行準(zhǔn)確的識(shí)別,使得導(dǎo)出文本出現(xiàn)嚴(yán)重錯(cuò)誤。2.3斷句例1:人工: a pentagon official said this was to

6、 provide president obama with flexibility.騰訊: a pentagon official said this was to provide president obama with flexibility should military options be required to protect american lives and interests.例2:人工:.people dont listen to them. why is that?搜狗:.people dont listen to them and why is that?騰訊:.pe

7、ople dont listen to them why is that?受原音頻語(yǔ)速和輕重讀音的影響,ai語(yǔ)音識(shí)別軟件難以像人工聽譯一樣做到準(zhǔn)確的斷句。但就普遍性而言,50個(gè)音頻里斷句錯(cuò)誤占比較低。絕大多數(shù)情況下,ai語(yǔ)音識(shí)別軟件還是能較準(zhǔn)確的對(duì)原音頻進(jìn)行斷句。2.4整體準(zhǔn)確性例1:人工:its the instrument we all play. its the most powerful sound in the world. probably its the only one that can start a war or say, i love you.訊飛:its the ins

8、trument we all play. probably see anyone that can start a war or say, i love you.搜狗:its the most powerful instrument well play. its the most powerful sound in the world. probably its the only one that can start a war or say, i love you.騰訊:voice instrument we will play its most powerful sound a world

9、 probably any one can start a war or say i love you.例2:人工:oh no, i cant leave you. i promised i would put your photo up. i promised you would see coco.訊飛:oh no, i cant leave you. i promised i put your photo up. i promise you would see coco.搜狗:its almost sunrise. leave you.騰訊:oh no, i cant leave you.

10、 i promised id put your phone up. i promised you would see coco.例3:人工:remember me though i have to say goodbye. dont let it make you cry. forever if im far away. look, i sing secret song to you. each time you hear sad guitar. know that im with you. the only way that i can be until youre in my arm ag

11、ain.訊飛:remember be so i have to travel for free man army each time you hear cent town with you noise to noise noise yeah yeah noise yeah.搜狗:remember be so i have to travel for free man army each time you hear cent town with you noise to noise yeah noise yeah.騰訊:real number me! do i have to say goodb

12、ye do not let it make you cry far away. i sings secret song to you. each time you hear sand it are. the only way that i can be until youre in my arm again.ai語(yǔ)音識(shí)別軟件在識(shí)別過(guò)程中,存在增聽、漏聽、連讀分辨不清、甚至部語(yǔ)段無(wú)法識(shí)別等問(wèn)題,使得識(shí)別后的文本正確率較源語(yǔ)文本低。人工聽譯主要依靠速記員的專業(yè)性,聽寫時(shí)長(zhǎng)長(zhǎng),且可反復(fù)聽寫某一模糊部分,正確率較源語(yǔ)文本高,準(zhǔn)確性較ai語(yǔ)音識(shí)別軟件更好。3.總結(jié)字幕聽譯較文本翻譯受到更多因素的限制。筆

13、者通過(guò)對(duì)人工聽譯與ai語(yǔ)音識(shí)別軟件聽譯的分析與對(duì)比發(fā)現(xiàn),人工能更好的保證斷句、口音校正和整體的準(zhǔn)確性,但用時(shí)長(zhǎng),工作量大,對(duì)速記員本身的語(yǔ)言素質(zhì)要求高;由于ai語(yǔ)音識(shí)別軟件當(dāng)前固有的問(wèn)題,ai語(yǔ)音識(shí)別整體上已經(jīng)達(dá)到不錯(cuò)水平,能較為準(zhǔn)確的識(shí)別出源音頻。這說(shuō)明,在日后的聽譯工作中,速記員可嘗試將ai語(yǔ)音識(shí)別后的文本作為藍(lán)本進(jìn)行再精聽;將ai語(yǔ)音識(shí)別技術(shù)同傳統(tǒng)聽譯結(jié)合起來(lái),采用更加靈活的聽譯策略和方法,更快速準(zhǔn)確的完成聽譯工作。參考文獻(xiàn)1林明月,耿磊.淺析字幕翻譯的特點(diǎn)j.明日風(fēng)尚,2016(18):282.2路雅芝.從功能對(duì)等理論淺談字幕聽譯以跨語(yǔ)言訪談?lì)惞?jié)目為例j.校園英語(yǔ),2019(14):229-230.3艾朝陽(yáng),周祎,李紅.

溫馨提示

  • 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。

最新文檔

評(píng)論

0/150

提交評(píng)論