音声領域におけるLINEの研究開発

Ի੠ྖҬʹ͓͚Δ -*/&ͷݚڀ։ൃ -*/&4QFFDI5FBN ໦ా ༞հ খদ ୡ໵

-*/&ͷ"*ٕज़ྖҬ LINE AI Speech Video Voice NLU Data OCR Vision
Face LINE Shopping Lens Adult Image Filter Scene Classification Ad image Filter Visual Search Analogous image Product Image Lip Reading Fashion Image Spot Clustering Food Image Indonesia LINE Split Bill LINE MUSIC Playlist OCR LINE CONOMI Handwritten Font Receipt OCR Credit card OCR Bill OCR Document Intelligence Identification Face Sign eKYC Face Sign Auto Cut Auto Cam Transcription Telephone network Voice recognition Single-Demand STT Simple voice High quality voice Voice Style Transfer Active Leaning Federated Leaning Action recognition Pose estimation Speech Note Vlive Auto Highlight Content Center AI CLOVA Dubbing LINE AiCall Gatebox Papago Video Insight LINE CLOVA AI Interactive Avatar Interactive Avatar Media 3D Avatar LINE Profile Lip Reading

-*/&ͷԻ੠ϓϩμΫτ LINE AiCall Call answering CLOVA Note Transcription

$-07"/PUF%FNP CLOVA Note Demo

5FDIOPMPHJFTJO$-07"/PUF End-to-End ASR based on Self-Supervised Learning Speaker Classification Keyword
Boosting

-*/&ͷԻ੠ٕज़ End-to-End / Hybrid ASR Text-to-Speech Acoustic Event Detection Multi-Channel
Speech Processing Research Product

ࠃࡍδϟʔφϧࠃࡍձٞ࠾୒࣮੷ 5PUBMBDDFQUBODFTJO Title Conf/Journal Author END TO END LEARNING FOR
CONVOLUTIVE MULTI-CHANNEL WIENER FILTERING ICASSP2021 M. Togami Disentangled speaker and language representations using mutual information minimization and domain adaptation for cross-lingual TTS ICASSP2021 T. Komatsu REFINEMENT OF DIRECTION OF ARRIVAL ESTIMATORS BY MAJORIZATION-MINIMIZATION OPTIMIZATION ON THE ARRAY MANIFOLD ICASSP2021 R. Scheibler SURROGATE SOURCE MODEL LEARNING FOR DETERMINED SOURCE SEPARATION ICASSP2021 R. Scheibler JOINT DEREVERBERATION AND SEPARATION WITH ITERATIVE SOURCE STEERING ICASSP2021 R. Scheibler, M. Togami PARALLEL WAVEFORM SYNTHESIS BASED ON GENERATIVE ADVERSARIAL NETWORKS WITH VOICING-AWARE CONDITIONAL DISCRIMINATORS ICASSP2021 R. Yamamoto TTS-BY-TTS: TTS-DRIVEN DATA AUGMENTATION FOR FAST AND HIGH-QUALITY SPEECH SYNTHESIS ICASSP2021 R. Yamamoto Independent Vector Analysis via Log-Quadratically Penalized Quadratic Minimization IEEE TSP R. Scheibler Multichannel Separation and Classification of Sound Events EUSIPCO2021 R. Scheibler, T. Komatsu, M. Togami Multi-Source Domain Adaptation with Sinkhorn Barycenter EUSIPCO2021 T. Komatsu Acoustic Event Detection with classifier chains INTERSPEECH2021 T. Komatsu Relaxing the Conditional Independence Assumption of CTC-based ASR by Conditioning on Intermediate Predictions INTERSPEECH2021 T. Komatsu Sound Source Localization with Majorization Minimization INTERSPEECH2021 M. Togami Efficient and Stable Adversarial Learning Using Unpaired Data for Unsupervised Multichannel Speech Separation INTERSPEECH2021 Y. Nakagome, M. Togami High-fidelity Parallel WaveGAN with Multi-band Harmonic-plus-Noise Model INTERSPEECH2021 R. Yamamoto Phrase break prediction with bidirectional encoder representations in Japanese text-to-speech synthesis INTERSPEECH2021 K. Futamata, B. Park, R. Yamamoto, K. Tachibana Over-Determined Semi-Blind Speech Source Separation APSIPA2021 M. Togami COMPARISON OF LOW COMPLEXITY SELF-ATTENTION MECHANISMS FOR ACOUSTIC EVENT DETECTION APSIPA2021 T. Komatsu, R. Scheibler A Comparative Study on Non-Autoregressive Modelings for Speech-to-Text Generation ASRU2021 T. Komatsu Computationally-Efficient Overdetermined Blind Source Separation Based on Iterative Source Steering IEEE SPL R. Scheibler

Ի੠ೝࣝ 4QFFDI3FDPHOJUJPO

Ի੠ೝࣝʹؔ͢Δ࠷ۙͷऔΓ૊Έ • ඇࣗݾճؼܕԻ੠ೝࣝʢ/POBVUPSFHSFTTJWF"43ʣ • $5$ʹجͮ͘/POBVUPSFHSFTTJWF"43 • ςΩετΛฒྻʹग़ྗ͢ΔͨΊߴ଎ • ฒྻग़ྗ ˠ
ςΩετؒͷґଘؔ܎ͷߟྀ͕ۤख • ߴ଎Ͱੑೳͷߴ͍$5$ʹجͮ͘ํࣜʹ͍ͭͯݚڀ • ೥Ҏ߱ͷऔΓ૊Έ • 3FMBYJOHUIF$POEJUJPOBM*OEFQFOEFODF"TTVNQUJPOPG$5$CBTFE"43 CZ$POEJUJPOJOHPO*OUFSNFEJBUF1SFEJDUJPOT </P[BLJBOE,PNBUTV *OUFSTQFFDI> • "$PNQBSBUJWF4UVEZPO/PO"VUPSFHSFTTJWF.PEFMJOHGPS4QFFDIUP5FYU(FOFSBUJPO <)JHVDIJ "436> • /PO"VUPSFHSFTTJWF"43XJUI4FMG$POEJUJPOFE'PMEFE&ODPEFST <,PNBUTV *$"441 BDDFQUFE > ͬͪ͜

3FMBYJOHUIF$POEJUJPOBM*OEFQFOEFODF"TTVNQUJPOPG$5$CBTFE"43 CZ$POEJUJPOJOHPO*OUFSNFEJBUF1SFEJDUJPOT </P[BLJBOE,PNBUTV *OUFSTQFFDI> 5SBOTGPSNFSͷதؒ૚Ͱ΋$5$ʹΑΔೝࣝΛߦ͍ޙஈͷϨΠϠʔ΁ϑΟʔυόοΫ 4FMGDPOEJUJPOFE$5$ தؒ૚ʹ͓͚Δೝࣝ݁ՌΛߟྀ͠ͳ͕Β࠷ऴͷೝࣝ݁ՌΛग़ྗ தؒ݁Ռ͕૚ΛܦΔ͝ͱʹʹྑ͘ͳͬͯΏ͘

"$PNQBSBUJWF4UVEZPO/PO"VUPSFHSFTTJWF.PEFMJOHGPS4QFFDIUP5FYU(FOFSBUJPO <)JHVDIJ "436> ૣҴాେ ඉޱ͞ΜʹΑΔൺֱ࿦จ <)JHVDIJ "436> 4FMGDPOEJUJPOFE$5$͕ෳ਺ͷ࠷৽ख๏ͷதͰτοϓͷੑೳΛୡ੒

/PO"VUPSFHSFTTJWF"43XJUI4FMG$POEJUJPOFE'PMEFE&ODPEFST <,PNBUTV *$"441 BDDFQUFE > 4FMGDPOEJUJPOFE$5$ͷܰྔԽʹؔ͢Δݚڀ 5SBOTGPSNFS $POGPSNFS ͷ֤૚͕ಉҰͷಇ͖Λ࣋ͭ͜ͱʹ஫໨ɺগ਺ϨΠϠʔΛ࠶ར༻͠ύϥϝʔλ࡟ݮ ಉҰͷઢܗ૚Ͱೝࣝ
ಛ௃ۭ͕ؒྨࣅ ύϥϝʔλ਺ Ͱैདྷͱಉ౳ͷੑೳΛୡ੒

ԻڹΠϕϯτݕग़ "DPVTUJD&WFOU%FUFDUJPO

ԻڹΠϕϯτݕग़ "DPVTUJD&WFOU%FUFDUJPO "&%ʣ • ԻڹΠϕϯτͷछྨɺൃੜ࣌ؒΛਪఆ͢ΔλεΫ .VMUJMBCFMDMBTTJGJDBUJPO • %$"4&8PSLTIPQ$IBMMFOHF͕ϕʔεϥΠϯ΍σʔληοτΛެ։ • %$"4&UBTLͰ͸
Ґ֫ಘ • $POWPMVUJPO"VHNFOUFE5SBOTGPSNFSGPS4FNJ4VQFSWJTFE4PVOE&WFOU%FUFDUJPO <.JZB[BLJ %$"4&> • ೥Ҏ߱ͷൃදจݙ • "DPVTUJD&WFOU%FUFDUJPOXJUI$MBTTJGJFS$IBJOT <,PNBUTV *OUFSTQFFDI> • $PNQBSJTPOPG-PX$PNQMFYJUZ4FMG"UUFOUJPO.FDIBOJTNTGPS"DPVTUJD&WFOU%FUFDUJPO <,PNBUTV "14*1"> • 4FMG4VQFSWJTFE-FBSOJOH.FUIPE6TJOH.VMUJQMF4BNQMJOH4USBUFHJFTGPS(FOFSBM1VSQPTF "VEJP3FQSFTFOUBUJPO <,VSPZBOBHJ *$"441 BDDFQUFE>

"DPVTUJD&WFOU%FUFDUJPOXJUI$MBTTJGJFS$IBJOT <,PNBUTV *OUFSTQFFDI> ैདྷͷԻڹΠϕϯτݕग़͸֤ΠϕϯτΛฒྻʹ෼ྨͦ͠ΕͧΕͷ૬ޓؔ܎Λߟྀͨ͠෼ྨ͕ۤख ԻڹΠϕϯτؒͷ૬ޓؔ܎Λߟྀͨ͠$IBJOSVMFʹجͮ͘ԻڹΠϕϯτ෼ྨثΛఏҊ ैདྷ֤ΠϕϯτΛͦΕͧΕ෼ྨ ઢܗ૚ 4JHNPJEʣ ఏҊ๏Πϕϯτ͝ͱʹࣗݾճؼతʹ෼ྨ ͭͷσʔληοτͰੑೳ޲্Λ֬ೝ
ಛʹ࣮ऩ࿥σʔλͰେ͖͘ੑೳվળ $MBTTJGJFS 'FBUVSF &YUSBDUPS *OQVU &WFOU" &WFOU# &WFOU$ 1SPQPTFE $MBTTJGJFS 'FBUVSF &YUSBDUPS *OQVU &WFOU" &WFOU# &WFOU$

$PNQBSJTPOPG-PX$PNQMFYJUZ4FMG"UUFOUJPO.FDIBOJTNTGPS"DPVTUJD&WFOU%FUFDUJPO <,PNBUTVBOE4DIFJCMFS "14*1"> "MMFYQFSJNFOUTBSFDPOEVDUFECBTFEPO5SBOTGPSNFSCBTFE"&%<.JZB[BLJ *$"441> • ԻڹΠϕϯτݕग़ʹ͓͚Δ5SBOTGPSNFS TFMGBUUFOUJPO ͱܥྻ௕ͷؔ܎ʹ͍ͭͯௐࠪ •
4FMGBUUFOUJPOͷܰྔԽʹؔ͢Δෳ਺ͷ࠷৽ख๏ͷੑೳධՁΛใࠂ

4FMG4VQFSWJTFE-FBSOJOH.FUIPE6TJOH.VMUJQMF4BNQMJOH4USBUFHJFTGPS(FOFSBM1VSQPTF "VEJP3FQSFTFOUBUJPO <,VSPZBOBHJBOE,PNBUTV *$"441 BDDFQUFE > • ࣗݾڭࢣ͋ΓֶशʹΑΔ൚༻తͳΦʔσΟΦදݱΛಘΔͨΊͷଛࣦؔ਺ʹؔ͢Δݚڀ • ࣗݾڭࢣ͋Γֶश
ˠ ଛࣦؔ਺ఆ͕ٛॏཁ • ैདྷ͸୯Ұ؍఺͔Βͷଛࣦؔ਺Λར༻ FH ಉҰͷԻϑΝΠϧʢdඵͳͲʣ಺ͷԻ͸͍ͣΕ΋ྨࣅͱԾఆ • ԾఆΛຬͨ͞ͳ͍λεΫ΋ଘࡏ • ఏҊख๏ෳ਺ͷ؍఺͔Βଛࣦؔ਺Λઃܭ • ˠϚϧνλεΫͳࣗݾڭࢣ͋Γֶशʹج͖ͮɺΑΓ൚༻తͳදݱΛ֫ಘ

ଟνϟϯωϧ৴߸ॲཧ .VMUJ$IBOOFM4JHOBM1SPDFTTJOH

4%3r .FEJVN3BSFXJUI'BTU$PNQVUBUJPOT<4DIFJCMFS> Goal: Accelerate Evaluation of Source Separation Algorithm MIX
SEPARATE … … … !" ̂ !$ !$ ̂ !" Faster Evaluation → More Experiments → Better Algorithms Reference Sources Estimated Sources Evaluation (SDR) We make this step faster!

4%3r .FEJVN3BSFXJUI'BTU$PNQVUBUJPOT<4DIFJCMFS> Signal-to-Distortion Ratio (SDR): • Similar to SNR
• Allow distortion by short filter ℎ (L=512) Filter ℎ is obtained by solving a linear system SDR = 10 log ℎ ⋆ , - ℎ ⋆ , − ̂ , - ,: reference ̂ ,: estimate min ℎ ⋆ , − ̂ , - Algorithm 1) Compute statistics 2) Solve for 3 3) Filter 4 4) Compute SDR Conventional 5(7 log 7) 5(9:) (Gaussian Elimination) 5(7 log 9) 5(7) Proposed 5(7 log 7) 5(9 log 9) (Conjugate Gradient Descent) - 5(9) (9 ≪ 7) 7: Number of samples 9: Filter length Significant savings because • 7 can be long! 30 sec. at 16 kHz is 480000 samples! • 9 is also large with 512 taps

4%3r .FEJVN3BSFXJUI'BTU$PNQVUBUJPOT<4DIFJCMFS> We release open-source implementation in Python with
numpy/torch backend pip install fast-bss-eval Benchmark vs open implementations • mir_eval • sigsep/bss_eval 8x 12x 15x

Ի੠߹੒ 5FYUUP4QFFDI

7PJDF5FBN Controllable, high-quality, and expressive TTS High-quality Neural Vocoder Controllable
TTS with emotion strength Text Analyzer

Controllable TTS with emotion strength Add emotion strength as the
one of input parameters on an acoustic model [ a- ri^ ga- to- o- ] ͋Γ͕ͱ͏ Acoustic model Text analyzer Vocoder Text &NPUJPO strength 0 1 Happiness Waveform

Prediction of emotion strength Consider two ways to predict emotion
strength from speech 8BWFGPSN Human annotators weak strong IBQQJOFTT weak strong 4BEOFTT Label emotion strength by listening Feature extractor Predict emotion strength with ranking algorithm J: Annotate automatically L: In some cases, decreasing expressiveness in TTS J: More expressive in TTS L: High cost

Acoustic model training Train acoustic model conditioning predicted emotion strength
and emotion label. Acoustic model and vocoder are trained independently. [ a- ri^ ga- to- o- ] ありがとう Text analyzer Acoustic model Vocoder Text Speech Predicting Emotion strength 0 1 Emotion strength Happiness Emotion strength and emotion label Speech

Speech samples Control emotion strength ! " )BQQZ 4BE #

ࠂ஌̍࠾༻৘ใ ˙۩ମతͳۀ຿಺༰ ɾԻ੠ॲཧٕज़ʹؔ͢Δݚڀ։ൃ ɾ&OEUP&OEܕԻ੠ೝࣝϞσϧɺσίʔμɺٴͼपลٕज़ ɾ%//)..ܕԻ੠ೝࣝϞσϧʢԻڹϞσϧɺݴޠϞσϧʣɺσίʔμɺٴͼपลٕज़ ɾ&OEUP&OEԻ੠߹੒ʹ޲͚ͨԻڹϞσϧɾOFVSBMWPDPEFS ɾԻڹΠϕϯτݕग़ɺϚϧνϞʔμϧॲཧͳͲ৽نϓϩμΫτʹܨ͕Δٕज़ ɾ࣍ͷϓϩμΫτͷछͱͳΔٕज़ͷߟҊɺٴͼϓϩτλΠϓγεςϜͷ։ൃ ɾαʔόӡ༻ɺσϓϩΠɺӡ༻؀ڥߏஙͱվળʢ.-0QTʣ ˙ϙδγϣϯͷັྗ
ɾ&OEUP&OEԻ੠ೝࣝ΍දݱྗʹண໨ͨ͠Ի੠߹੒ΛؚΉ࠷ઌ୺ͷٕज़ͷϓϩμΫτԽʹܞΘΔ͜ͱ͕Ͱ͖·͢ ɾ-*/&ΞϓϦ΍εϚʔτεϐʔΧʔ౳ͷطଘϓϩμΫτΛ௨ͯ͠େن໛ͳϢʔβʔͷੜ׆ʹӨڹΛ༩͑ΒΕΔͱ ಉ࣌ʹɺ-*/&"J$BMMͷΑ͏ʹͦΕ·Ͱੈʹͳ͔ͬͨ৽͍͠ϓϩμΫτͷ্ཱͪ͛ʹࢀը͢Δػձ͕͋Γ·͢ ɾ*$"441ɺ*/5&341&&$)ͱ͍ͬͨτοϓࠃࡍձٞʹͯൃදܦݧΛ࣋ͭଟ͘ͷΤϯδχΞɾϦαʔνϟʔͱ ڞʹݚڀ։ൃ͕Ͱ͖Δ؀ڥͰ͢ ˙։ൃ؀ڥ ɾ։ൃݴޠɿ1ZUIPO $$ 4IFMM4DSJQUͳͲ ɾ04ɿ-JOVY .BD049 ɾ෼ࢄॲཧɿ"QBDIF4QBSL )BEPPQ ɾίϯςφɿ%PDLFS ,VCFSOFUFT ɾ$*$%ɿ%SPOF "SHP +FOLJOT ɾ.-ϥΠϒϥϦɿ1Z5PSDI 5FOTPS'MPXͳͲ ɾͦͷଞɿ(JU)VC&OUFSQSJTF $POGMVFODF 4MBDL

ࠂ஌̎ ೥ՆͷΠϯλʔϯʢ༧ఆʣ Ի੠ٕज़ʹؔΘΔ࣮՝୊ɾλεΫͷʹऔΓ૊Ή։ൃ ݚڀΠϯλʔϯΛืू༧ఆͰ͢ɻ ظؒɿ݄d݄ͷؒͰिؒʢ౔೔ॕٳΈʣ څ༩ࢧڅɿສԁʢສԁिʣ ۈ຿஍ɿΦϯϥΠϯ ςʔϚͷྫɿ ɾ&&Ի੠ೝࣝɺԻ੠߹੒ ɾ؀ڥԻ෼ੳ
ɾϚϧνϞʔμϧ FHBVEJPWJTVBM"43ʣ ΤϯτϦʔαΠτ͸݄Լ०ʹΦʔϓϯ༧ఆͰ͢ɻΤϯτϦʔ͓଴͍ͪͯ͠·͢ɻ

ࠂ஌̏ We Have a Presentation at NVIDIA GTC2022!! Title: Building
Streaming End-to-end Speech Recognition Service with Triton Inference Server

音声領域におけるLINEの研究開発

音声領域におけるLINEの研究開発

LINE Developers
PRO

More Decks by LINE Developers

Other Decks in Technology

Featured

Transcript

Ի੠ྖҬʹ͓͚Δ -/&ͷݚڀ։ൃ -/&4QFFDI5FBN ໦ా ༞հ খদ ୡ໵

-/&ͷ"ٕज़ྖҬ LINE AI Speech Video Voice NLU Data OCR Vision

-*/&ͷԻ੠ϓϩμΫτ LINE AiCall Call answering CLOVA Note Transcription

$-07"/PUF%FNP CLOVA Note Demo

5FDIOPMPHJFTJO$-07"/PUF End-to-End ASR based on Self-Supervised Learning Speaker Classification Keyword

-*/&ͷԻ੠ٕज़ End-to-End / Hybrid ASR Text-to-Speech Acoustic Event Detection Multi-Channel

ࠃࡍδϟʔφϧࠃࡍձٞ࠾୒࣮੷ 5PUBMBDDFQUBODFTJO Title Conf/Journal Author END TO END LEARNING FOR

Ի੠ೝࣝ 4QFFDI3FDPHOJUJPO

Ի੠ೝࣝʹؔ͢Δ࠷ۙͷऔΓ૊Έ • ඇࣗݾճؼܕԻ੠ೝࣝʢ/POBVUPSFHSFTTJWF"43ʣ • $5$ʹجͮ͘/POBVUPSFHSFTTJWF"43 • ςΩετΛฒྻʹग़ྗ͢ΔͨΊߴ଎ • ฒྻग़ྗ ˠ

"$PNQBSBUJWF4UVEZPO/PO"VUPSFHSFTTJWF.PEFMJOHGPS4QFFDIUP5FYU(FOFSBUJPO <)JHVDIJ "436> ૣҴాେ ඉޱ͞ΜʹΑΔൺֱ࿦จ <)JHVDIJ "436> 4FMGDPOEJUJPOFE$5$͕ෳ਺ͷ࠷৽ख๏ͷதͰτοϓͷੑೳΛୡ੒

/PO"VUPSFHSFTTJWF"43XJUI4FMG$POEJUJPOFE'PMEFE&ODPEFST <,PNBUTV *$"441 BDDFQUFE > 4FMGDPOEJUJPOFE$5$ͷܰྔԽʹؔ͢Δݚڀ 5SBOTGPSNFS $POGPSNFS ͷ֤૚͕ಉҰͷಇ͖Λ࣋ͭ͜ͱʹ஫໨ɺগ਺ϨΠϠʔΛ࠶ར༻͠ύϥϝʔλ࡟ݮ ಉҰͷઢܗ૚Ͱೝࣝ

ԻڹΠϕϯτݕग़ "DPVTUJD&WFOU%FUFDUJPO

ԻڹΠϕϯτݕग़ "DPVTUJD&WFOU%FUFDUJPO "&%ʣ • ԻڹΠϕϯτͷछྨɺൃੜ࣌ؒΛਪఆ͢ΔλεΫ .VMUJMBCFMDMBTTJGJDBUJPO • %$"4&8PSLTIPQ$IBMMFOHF͕ϕʔεϥΠϯ΍σʔληοτΛެ։ • %$"4&UBTLͰ͸

$PNQBSJTPOPG-PX$PNQMFYJUZ4FMG"UUFOUJPO.FDIBOJTNTGPS"DPVTUJD&WFOU%FUFDUJPO <,PNBUTVBOE4DIFJCMFS "141"> "MMFYQFSJNFOUTBSFDPOEVDUFECBTFEPO5SBOTGPSNFSCBTFE"&%<.JZB[BLJ $"441> • ԻڹΠϕϯτݕग़ʹ͓͚Δ5SBOTGPSNFS TFMGBUUFOUJPO ͱܥྻ௕ͷؔ܎ʹ͍ͭͯௐࠪ •

4FMG4VQFSWJTFE-FBSOJOH.FUIPE6TJOH.VMUJQMF4BNQMJOH4USBUFHJFTGPS(FOFSBM1VSQPTF "VEJP3FQSFTFOUBUJPO <,VSPZBOBHJBOE,PNBUTV *$"441 BDDFQUFE > • ࣗݾڭࢣ͋ΓֶशʹΑΔ൚༻తͳΦʔσΟΦදݱΛಘΔͨΊͷଛࣦؔ਺ʹؔ͢Δݚڀ • ࣗݾڭࢣ͋Γֶश

ଟνϟϯωϧ৴߸ॲཧ .VMUJ$IBOOFM4JHOBM1SPDFTTJOH

4%3r .FEJVN3BSFXJUI'BTU$PNQVUBUJPOT<4DIFJCMFS> Goal: Accelerate Evaluation of Source Separation Algorithm MIX

4%3r .FEJVN3BSFXJUI'BTU$PNQVUBUJPOT<4DIFJCMFS> Signal-to-Distortion Ratio (SDR): • Similar to SNR

4%3r .FEJVN3BSFXJUI'BTU$PNQVUBUJPOT<4DIFJCMFS> We release open-source implementation in Python with

Ի੠߹੒ 5FYUUP4QFFDI

7PJDF5FBN Controllable, high-quality, and expressive TTS High-quality Neural Vocoder Controllable

Controllable TTS with emotion strength Add emotion strength as the

Prediction of emotion strength Consider two ways to predict emotion

Acoustic model training Train acoustic model conditioning predicted emotion strength

Speech samples Control emotion strength ! " )BQQZ 4BE #

ࠂ஌̎ ೥ՆͷΠϯλʔϯʢ༧ఆʣ Ի੠ٕज़ʹؔΘΔ࣮՝୊ɾλεΫͷʹऔΓ૊Ή։ൃ ݚڀΠϯλʔϯΛืू༧ఆͰ͢ɻ ظؒɿ݄d݄ͷؒͰिؒʢ౔೔ॕٳΈʣ څ༩ࢧڅɿສԁʢສԁिʣ ۈ຿஍ɿΦϯϥΠϯ ςʔϚͷྫɿ ɾ&&Ի੠ೝࣝɺԻ੠߹੒ ɾ؀ڥԻ෼ੳ

ࠂ஌̏ We Have a Presentation at NVIDIA GTC2022!! Title: Building