Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
音声情報処理に便利な (Python) パッケージやソフトウェア
Search
Sponsored
·
Your Podcast. Everywhere. Effortlessly.
Share. Educate. Inspire. Entertain. You do you. We'll handle the rest.
→
Akira Tamamori
December 30, 2020
Research
960
3
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
音声情報処理に便利な (Python) パッケージやソフトウェア
Tokyo BISH Bashの資料から独立させたもの
Akira Tamamori
December 30, 2020
More Decks by Akira Tamamori
See All by Akira Tamamori
音声認識と音声合成の超入門
tam17aki
0
540
Tokyo BISH Bash #02 音声情報処理と音声変換技術入門
tam17aki
2
2.3k
[ICASSP2020音響音声読み会] State-Space Gaussian Process for Drift Estimation in Stochastic Differential Equations
tam17aki
0
590
Other Decks in Research
See All in Research
第64回CV・PRML勉強会 論文紹介:Linguistic Priors for Visual Decoupling: Towards Symmetric Vision-Brain Alignment
sokikatayama
0
150
羽田新ルート運用6年の検証
1manken
0
180
LINEヤフー データサイエンス Meetup「三井物産コモディティ予測チャレンジ」の舞台裏-AlpacaTechパート
gamella
1
620
AIで最適化を解けるか?
mickey_kubo
0
150
医療LLMの現在地〜最新研究から社会実装までを考える〜
kento1109
1
1.6k
Apache Gravitinoで実現する Icebergカタログ統合とアクセスの一元化
matsumooon
0
400
[IR Reading 2026春 論文紹介] LLM-based Listwise Reranking under the Effect of Positional Bias (ECIR 2026) /IR-Reading-2026-Spring
koheishinden
PRO
0
300
National high-resolution cropland classification of Japan with agricultural census information and multi-temporal multi-modality datasets
satai
3
410
HackSick vol.7 LT資料【LLMアーキテクチャ入門・事前学習時の躓き所解説】 スパースなAttention・状態空間モデル
rikkabotan7
0
140
Dual Quadric表現を用いた動的物体追跡とRGB-D・IMU制約の密結合によるオドメトリ推定
nanoshimarobot
0
490
MM-OVSeg: Multimodal Optical–SAR Fusion for Open-Vocabulary Segmentation in Remote Sensing
satai
3
110
Harness Engineering and Al Agent
kzinmr
3
1.9k
Featured
See All Featured
Breaking role norms: Why Content Design is so much more than writing copy - Taylor Woolridge
uxyall
0
360
Connecting the Dots Between Site Speed, User Experience & Your Business [WebExpo 2025]
tammyeverts
11
980
Exploring the Power of Turbo Streams & Action Cable | RailsConf2023
kevinliebholz
37
6.5k
16th Malabo Montpellier Forum Presentation
akademiya2063
PRO
0
330
Leveraging Curiosity to Care for An Aging Population
cassininazir
1
460
Building a Modern Day E-commerce SEO Strategy
aleyda
45
9.2k
[Rails World 2023 - Day 1 Closing Keynote] - The Magic of Rails
eileencodes
38
2.9k
Paper Plane
katiecoart
PRO
2
52k
The Impact of AI in SEO - AI Overviews June 2024 Edition
aleyda
5
1.2k
How to Think Like a Performance Engineer
csswizardry
28
2.7k
10 Git Anti Patterns You Should be Aware of
lemiorhan
PRO
659
62k
The AI Revolution Will Not Be Monopolized: How open-source beats economies of scale, even for LLMs
inesmontani
PRO
3
3.7k
Transcript
ԻใॲཧԻมʹ ศརͳ (Python) ύοέʔδ ιϑτΣΞͨͪ 1 ༻ײͳͲࢲݟΛؚΈ·͢
TPYQZTPY • ίϚϯυϥΠϯ͔ΒϑΥʔϚοτมͳͲΛ͓खܰʹ • ϑΥʔϚοτมʢwav to mp3 ͳͲʣ • ݁߹ϛοΫεɺτϦϛϯά
Մೳ • όονॲཧָʢγΣϧεΫϦϓτͳͲʣ • Pythonϥούʔ pysox͋Δ • Πϯετʔϧ • brew install sox ͳͲ • pip install sox ← pysox ͷΠϯετʔϧ͜Ε 2
MJCSPTBʢ͜ΕຊʹΦεεϝʣ • Ի/ԻָͷੳʹศརͳϞδϡʔϧ͕ଗͬͨύοέʔδ • ެࣜϚχϡΞϧɾνϡʔτϦΞϧͷॆ࣮ॿ͔Δ • ݸਓతʹΑ͘͏ػೳ • ܗදࣔɺεϖΫτϩάϥϜදࣔ •
Իಛྔநग़ʢରϝϧεϖΫτϩάϥϜʣ • Πϯετʔϧ pip install librosa • ެࣜϖʔδ https://librosa.org/librosa/index.html 3
1Z8PSME • Իͷੳ࠶߹Λߦ͏Ϙίʔμʔͷύοέʔδ • ԻΛʮ৭ɾͷߴ͞ɾͷ͔͢Εʯͷ֤ʹղ͠࠶߹ • C++൛ͷPythonϥούʔ • Իͷಛྔநग़ʹ͑ͯศར ⇒
PySPTKʢޙड़ʣΑΓ࣭ͷΑ͍εϖΫτϧแབྷ • Πϯετʔϧ pip install pyworld • ެࣜϖʔδ https://github.com/JeremyCCHsu/Python-Wrapper-for-World-Vocoder 4
1Z"VEJP • ετϦʔϜԻ / ࠶ੜʹศརͳύοέʔδ • ϦΞϧλΠϜͷԻೖྗɾԻग़ྗʹ͑Δ • ϦΞϧλΠϜԻม with
PythonͳͲՄೳ • Πϯετʔϧ • pip install pyaudio ※ཁportaudio (e.g., brew install portaudio) 5
1Z"VEJPͱ1Z8PSMEͷΈ߹Θͤ • ؆қ൛ͷϘΠενΣϯδϟʔ • ؆қϘΠενΣϯδϟʔͷεΫϦϓτΛվྑɿPyQt5ͷεϥΠμʔʹΑΓ ϐονͱϑΥϧϚϯτΛϦΞϧλΠϜௐ͢ΔػೳΛՃʢฐϒϩάʣ • banibiku • Zoom৴͚ʹ̎࣍ݩΩϟϥʹͳΓ͖Δ͜ͱΛࢦͨ͠ϓϩδΣΫτ
• scripts/voice_converter.py ͕ྑ͍ײ͡ͷϘΠενΣϯδϟʔ → ฐϒϩάͷαϯϓϧεΫϦϓτͷόάϑΟοΫεؚ͕·ΕΔ 6 https://tam5917.hatenablog.com/entry/2019/04/30/213321 https://github.com/peisuke/babiniku
1Z415, • ԻใॲཧπʔϧΩοτSPTKͷPythonϥούʔ • SPTKࣗମLinuxίϚϯυ܈ • Իڹಛྔநग़ʹ͏ͷ͕ศར • Իੳ߹Ͱ͖Δ͕ɺ࣭ࣗମWORLDͷ΄͏্͕ •
Πϯετʔϧ pip install pysptk • ެࣜϖʔδ https://pysptk.readthedocs.io/en/latest/ 7
OONOLXJJ <OBOBNJO LBXBJJ> • DNNԻ߹ʹཱͭϞδϡʔϧΛूΊͨύοέʔδ • ͲͪΒ͔ͱ͍͏ͱݚڀ༻్ • લॲཧԻڹಛྔநग़ͷΫϥε͕Ұ௨Γଗ͍ͬͯΔ •
จͷ࠶ݱ࣮Λ͢Δͱ͖ͳͲʹେ͍ʹཱͭ • Πϯετʔϧ pip install nnmnkwii • ެࣜϖʔδ https://r9y9.github.io/nnmnkwii/stable/index.html 8
1ZEVC • Pydub • ܗฤूʹศརͳϞδϡʔϧΛूΊͨύοέʔδ • αϙʔτ͢ΔϑΝΠϧܗࣜ๛ʢwav, mp3, mp4, wma,
aac, ...ʣ • ػೳ Γग़͠ɺׂɺϛοΫεɺϑΣʔυΠϯΞτɺແԻૠೖɺͳͲͳͲ • Ұ෦ͷػೳ pysoxͷ΄͏͕ߴͱ͍͏ӟ?ʢະ֬ೝʣ • Πϯετʔϧ pip install pydub • ެࣜϖʔδ http://pydub.com/ 9
TQSPDLFU • ౷ܭత࣭มͷͨΊͷπʔϧΩοτ (not ύοέʔδ) • ͲͪΒ͔ͱ͍͏ͱݚڀ༻ ʢMITϥΠηϯεʣ • ݚڀͷʮϕʔεϥΠϯʯߏஙʹ࠷ద
• ެࣜϖʔδ https://github.com/k2kobayashi/sprocket • ղઆจ ʰ౷ܭత࣭มιϑτΣΞೖʱ https://www.jstage.jst.go.jp/article/isciesci/62/2/62_69/_article/-char/ja/ • νϡʔτϦΞϧ (εϥΠυ & notebook) https://github.com/kan-bayashi/INTERSPEECH19_TUTORIAL 10
"VEBDJUZ ೖΕ͓ͯ͘ͱ҆৺ • ϑϦʔͷܗฤूιϑτɺϚϧνϓϥοτϑΥʔϜ • ๛ͳαϯυΤϑΣΫτՃػೳ • ެࣜϖʔδ https://www.audacityteam.org/ 11
(16্ͰԻॲཧ͍ͨ͠Ϛϯʹ ͓͢͢Ίͷύοέʔδ 12 ͓·͚
UPSDIBVEJP • Pytorchެ͕ࣜαϙʔτ͍ͯ͠ΔԻॲཧܥϥΠϒϥϦ • PytorchܥͷਂֶशϞσϧͱͷ૬ੑ͕ྑ͍ʢͦΕͦ͏ʣ • ެࣜϖʔδ https://pytorch.org/audio/stable/index.html 13
UGTJHOBM • TensorFlowެ͕ࣜαϙʔτ͍ͯ͠ΔԻॲཧܥͷؔ܈ • TFܥͷਂֶशϞσϧͱͷ૬ੑ͕ྑ͍ʢͦΕͦ͏ʣ • FFT/iFFT, DCT, MDCT, STFTͳͲ
• ެࣜϖʔδ https://www.tensorflow.org/api_docs/python/tf/signal 14
UPSDIMJCSPTB • PytorchΛόοΫΤϯυʹͯ͠librosaΛGPU্Ͱಈ͔͢ • Πϯετʔϧ pip install torchlibrosa • ެࣜϖʔδ
https://github.com/qiuqiangkong/torchlibrosa 15
LBQSF • Kerasʢͱ͍͏͔TFʣΛόοΫΤϯυʹͯ͠Իॲཧ͢Δ • STFTiSTFTɺϝϧεϖΫτϩάϥϜͳͲ • CQTͳͲͳ͍ • ެࣜϖʔδ https://github.com/keunwoochoi/kapre
• Πϯετʔϧ pip install kapre 16
OO"VEJP • PytorchΛόοΫΤϯυʹͯ͠STFTͳͲΛGPU্Ͱಈ͔͢ • STFTɺٯSTFTɺCQTͳͲΑ͘͏ಛநग़ܥ͕ଗ͏ • ެࣜϖʔδ https://github.com/KinWaiCheuk/nnAudio 17
OO"VEJPʢ͖ͭͮʣ • ൺֱද 18