Speech to Text to Speech

/ home / spielereien / .

Siehe auch: Artificial Intelligence (AI)

Speech to Text

• Kaldi ASR:
  • https://sourceforge.net/p/kaldi/code/HEAD/tree/sandbox/idlak/
  • https://github.com/kaldi-asr/kaldi
  • https://kaldi-asr.org/
  • https://kaldi-asr.org/doc/kaldi_for_dummies.html
  • https://kaldi-asr.org/models.html
    1. Package -> no package available!!!
    2. Compile (Ubuntu 18.04):
       cd /tmp/build
git clone https://github.com/kaldi-asr/kaldi.git kaldi --origin upstream
cd kaldi/tools
extras/check_dependencies.sh
make -j4

extras/install_irstlm.sh
cd ..
source tools/env.sh
cd ..

cd kaldi/src
./configure --shared
make -j4 depend
make -j4
cd ../..
       In order to install the language model kaldi_lm: cd kaldi/tools
extras/install_kaldi_lm.sh
       
• Vosk

Speech to Text on Debian

• voice2json
  • https://github.com/synesthesiam/voice2json
  • https://voice2json.org/commands.html
• https://blends.debian.org/accessibility/tasks/speechrecognition
• https://www.debian.org/devel/debian-accessibility/software
• Pocketsphinx
  $ pocketsphinx_continuous -infile audio.wav
-mdef option or with -hmm
  • https://cmusphinx.github.io/
  • https://github.com/cmusphinx/pocketsphinx
  • https://confusedcoders.com/random/speech-recognition-in-python-with-cmu-pocketsphinx
  • https://www.neurobs.com/pres_docs/html/03_presentation/05_response_devices/03_speech_recognition/01_speech_recognition_setup/01_sphinx_recognition_engine.htm

Extract sound from video

See also: YouTube Transkript erstellen

$ VIDEO="tgws21.mp4"
$ AUDIO="audio.wav"

$ ffmpeg -i ${VIDEO} ${AUDIO}

$ ffmpeg -i ${VIDEO} -vn -acodec libvorbis audio.ogg

• https://askubuntu.com/questions/41749/how-can-i-extract-audio-track-from-video-file-and-save-it-as-mp3
• https://linuxconfig.org/using-ffmpeg-to-extract-audio-from-mp4-media-file-on-linux

Open Source Speech Recognition Software Solutions

• Simon http://simon-listens.blogspot.com/2012/12/simon-040.html
• Kaldi
• CMU-Sphinx Speech Recognition http://jrmeyer.github.io/asr/2016/01/09/Installing-CMU-Sphinx-on-Ubuntu.html http://cmusphinx.sourceforge.net/
• Mozilla
• Julius https://github.com/julius-speech/julius
• Dictation Bridge
• Mycroft
• Dragon NaturallySpeaking
• HTK
• Voxforge http://voxforge.org/home/dev

Quelle: https://www.goodfirms.co/speech-recognition-software/blog/best-free-open-source-speech-recognition-software

• DeepSpeech https://github.com/mozilla/DeepSpeech
• Kaldi http://kaldi.sourceforge.net/ https://github.com/alumae/kaldi-gstreamer-server
• Julius http://julius.osdn.jp/
• Wav2Letter++ (Facebook) https://github.com/flashlight/wav2letter -> Flashlight https://github.com/flashlight/flashlight
• DeepSpeech2
• Open Seq2Seq
• Fairseq
• Vosk (deutsch, englisch) https://alphacephei.com/vosk/
• Athena
• ESPnet
• Simon
• Eesen
• deepspeech.pytorch
• CVoiceControl http://www.kiecza.net/daniel/linux/
• Ears http://www.speech.cs.cmu.edu/comp.speech/Section6/Recognition/ears.html

Quellen:

• https://unix.stackexchange.com/questions/256138/is-there-any-decent-speech-recognition-software-for-linux
• https://fosspost.org/open-source-speech-recognition/

2021/2022:

• DeepSpeech (Mozilla) -> Google Tensor Flow???, Latest Release 2020-12 https://deepspeech.readthedocs.io/en/r0.9/?badge=latest
• Wav2Letter++
• Kaldi
• Open Seq2Seq
• Tensorflow ASR (Google) https://www.tensorflow.org/
• https://www.rev.com/blog/resources/the-5-best-open-source-speech-recognition-engines-apis
• Amberscript (39 languages)
• Trint
• Rev
• Sonix
• Veed
• 360converter
• Type Studio
• Go Transcribe
• Happy Scribe
• https://geekflare.com/transcribe-video-to-text/
• Vosk https://github.com/alphacep/vosk-api
• MaryTTS https://github.com/marytts http://mary.dfki.de/
• HTK https://htk.eng.cam.ac.uk/download.shtml
• Dragon: https://www.nuance.com/dragon.html
• oTranscribe https://otranscribe.com/opensource/
• DeepSpeech
• Kaldi
• Julius
• Wav2Letter++
• DeepSpeech2
• Open Seq2Seq
• Fairseq
• Vosk
• Athena
• ESPnet
• https://gnulinux.ch/open-source-spracherkennung
• eSpeak http://espeak.sourceforge.net/
• MaryTTS
• Text to Speech Converter von avi-jkiapt
• Textkonverter
• Online-TTS
• Text-to-Speech
• TextToSpeechConverter
• https://br.atsit.in/de/?p=68224

21 Beste Online-Software für die Sprachausgabe

• https://text2audio.com/de/
• https://blog.codecentric.de/vergleich-von-text-to-speech-systemen

OCR

• Tesseract-OCR: https://github.com/tesseract-ocr/tesseract (12 Sprachen)
• EasyOCR
• PaddleOCR
• GOCR (2 Sprachen)
• Cognitive OpenOCR
• MMOCR
• KerasOCR
• https://www.hellosign.com/blog/best-free-and-open-source-optical-character-recognition-ocr-software
• https://medium.com/technovators/review-of-best-open-source-ocr-tools-fc839a20e61f
• Calamari https://github.com/Calamari-OCR/calamari (Tensorflow)
• OCRopus
• Kraken
• Tesseract
• TrOCR
• https://source.opennews.org/articles/so-many-ocr-options/
• Mit dem SSML-Tags-Editor können Sie sogar Faktoren wie Aussprache, Lautstärke und Sprachrate steuern.
• https://towardsdatascience.com/5-open-source-tools-you-can-use-to-train-and-deploy-an-ocr-project-8f204dec862b
• FreeOCR
• CVision OCR Engine
• OCR.Space

Quelle: https://www.goodfirms.co/ocr-software/blog/best-free-open-source-ocr-software

Textkorrektur

• LanguageTool https://github.com/Erikvl87/docker-languagetool https://languagetool.org/de
• Solr: https://solr.apache.org/

Audio to text

• Ferdinand, Linux News, 2026-01-09: EasySpeak: Sprachsteuerung für den Linux-Desktop
• DeepSpeech documentation
• GitHub: Mozilla DeepSpeech
• Yujian Tang, 13. Oktober 2021: DeepSpeech for Dummies - A Tutorial and Overview
• Immo Junghärtchen, c’t Magazin / heise.de, 26. Mai 2023: Spracherkennung und Transkription mit KI: Sprache in Text umwandeln mit Whisper - Die Open-Source-Spracherkennung Whisper transkribiert Sprache aus Audiodateien mit sehr guter Erkennungsquote und versteht sich sogar auf Zeichensetzung.
• Ralf Hersel, GNU/Linux.ch, 24. August 2020: Open Source Spracherkennung - Ein Überblick über aktuelle Speech-to-Text Systeme
  • DeepSpeech, Kaldi, Julius, Wav2Letter++, DeepSpeech2, OpenSeq2Seq, Fairseq, Vosk, Athena, ESPnet
• Vanessa Arnold, neuroflash, 10. Juli 2023: Whisper OpenAI: Sprache in Text umwandeln wie ein Profi
• Kaldi: https://kaldi-asr.org/
  https://github.com/kaldi-asr/kaldi
• Julius: dead? Packages!
  julius.osdn.jp/en_index.php
  https://github.com/julius-speech/julius
  https://github.com/julius-speech/julius/blob/master/Sample.jconf
  CLI
  https://www.voxforge.org/home/dev/acousticmodels/linux/create/htkjulius/tutorial/run-julius
  https://oceanai.mit.edu/pavlab/pdfs/app_uspeech_rec.pdf
• wav2letter / Wav2Letter++ / –> Flashlight ASR
  https://github.com/flashlight/wav2letter
• DeepSpeech2, PaddleSpeech
• OpenSeq2Seq
• Fairseq
• Vosk
• Athena
• ESPnet / espnet
• KDE Simon, no packages!
  https://simon.kde.org/
  https://speechify.com/product-reviews/simon-speech-recognition/?landing_url=https%3A%2F%2Fspeechify.com%2Fproduct-reviews%2Fsimon-speech-recognition%2F
  https://github.com/KDE/simon
• Kdenlive (based on Vosk), packages are there!
  https://docs.kdenlive.org/en/effects_and_compositions/speech_to_text.html
  -> Overcomplicated!!!
• KMouth, text to speach!!!
• Gnome Orca, screen reader: https://help.gnome.org/users/orca/stable/introduction.html.en
• Speech Recognition Software: https://tldp.org/HOWTO/Speech-Recognition-HOWTO/software.html
• https://en.wikipedia.org/wiki/Speech_recognition_software_for_Linux
• https://github.com/flashlight/flashlight
• https://github.com/cmirnow/Google-Cloud-TTS-Rails
• https://unix.stackexchange.com/questions/256138/is-there-any-decent-speech-recognition-software-for-linux
• https://www.goodfirms.co/speech-recognition-software/blog/best-free-open-source-speech-recognition-software
• https://fosspost.org/open-source-speech-recognition/