[Перевод] OpenAI взимает оплату поминутно, так что сделаем минуты короче

Я имею в виду буквально. Перед транскрибацией увеличьте скорость аудио в два или в три раза в ffmpeg . Благодаря этому вы потратите меньше токенов и вам придётся меньше ждать, а качество транскрипции почти не снизится. Вот и всё! Ниже показан скрипт, сочетающий все мои трюки для выполнения этой задачи. Для его работы необходимо установить yt-dlp , ffmpeg и llm . Я сэкономил вам время, сразу перейдя к делу, но если вы хотите подробностей истории о том, как я случайно обнаружил этот способ, пытаясь создать краткое изложение сорокаминутного доклада Андрея Карпаты, то продолжайте чтение. Вам стоит продолжить чтение, если вы недоумеваете, почему я просто не воспользовался встроенной автотранскрибацией YouTube. Краткий ответ будет таким: я недотёпа, считавший (ошибочно), что она недоступна. Поэтому мне пришлось идти по более сложному пути.

https://habr.com/ru/articles/922070/

#openai #транскрибация #транскрипция #voicetotext

Latest #ChatGPT update on Android removes speech timer, making #VoiceToText even worse. This keyboard app with only 1k users and no reviews is an ideal fix. Faster than #Whisper. No time limit. 10mins speech converts in <5 seconds! Works in any app. Uses #Grok API. $3 lifetime, or #OpenSource on #Fdroid (with limitations)
https://play.google.com/store/apps/details?id=net.devemperor.dictate

Come dettare su Word un testo o un documento
#Accessibilità #Android #Consigli #Dettatura #Guida #iOS #iPadOS #macOS #Microsoft #Microsoft365 #MicrosoftWord #Produttività #Suggerimenti #Tecnologia #Tutorial #VoiceToText #Windows #Word

https://www.ceotech.it/come-dettare-su-word-un-testo-o-un-documento/

Omi is what happens when wearable tech meets AI.
Put it on. Speak. It records — automatically.
No buttons, no hassle. Just seamless memory.

GitHub: github.com/BasedHardware/omi

#WearableAI #OpenSourceHardware #VoiceToText #Lifelogging #AIOnTheGo #Omi

✔️ Voice-to-Text Transcription Bot for Paid Telegram Groups

✨A #TelegramBot that auto-transcribes all voice messages inside paid or private groups and delivers searchable text to subscribers—ideal for busy users or multilingual audiences.

#VoicetoText #Transcription #GroupProductivity #ContentAccessibility #BusyProfessionals

@farooqkz a pro tip. If you use voice to text it always makes Linux “Lennox” but if you add “Linux” as a contact in your phone it always gets it right!
#phone #voicetotext #linux

📝 Turn your voice notes into organized Notion docs effortlessly with Notis AI! 📱 Simply send your voice messages via WhatsApp and let AI transcribe & structure everything in your Notion workspace. Perfect for busy professionals! #Notion #AI #VoiceToText #Productivity

https://pupuweb.com/how-to-turn-voice-notes-into-notion-docs-with-notis-ai/

➡️ Full code on GitHub: https://github.com/shinglyu/WhisperNow

#Linux #Transcription #FasterWhisper #OpenSource #Productivity #DIY #TechBlog #VoiceToText

Need quick audio transcription? Just Press Record turns your voice into text instantly across devices! Great for notes, interviews, & more. #Productivity #VoiceToText

WhatsApp introduces voice note transcriptions, offering on-device privacy and accessibility for converting audio to text. #ai #meta #whatsapp #voicetotext

https://buff.ly/3AFyyTO

I haven't been paying close attention to Whisper of late, but this is the first I've heard of this issue, though I guess I shouldn't be as surprised as I am 😑

OpenAI's Whisper invents parts of transcriptions — a lot https://www.engadget.com/ai/openais-whisper-invents-parts-of-transcriptions--a-lot-120039028.html?src=rss

#transcription #AI #Whisper #VoiceToText

After having a “conversation” in the Group chat today with multiple 6-9 min long as voice messages. I cant wait for apple to finally introduce voice to text as shown in previous demos. As I would rather read 2 pages of text or have a fucking phone call than to listen to 15min of someone sending me voice messages.

Don’t get me wrong I love my friends but I hate listening to those messages.

#apple #voicetotext #messages #chat #rant

Nanosecond (noun): The amount of time that lapses between hitting SEND and recognizing what autofill and voice-to-text did to your words.

#technology #texting #VoiceToText #time #autofill

I need some recommendations for an easy to use #VoiceToText physical product or app, please & thank you. It's for my creative writing projects.

I'm more of an oral storyteller & find it difficult to just write. I'm used to just speaking & instead of trying to force my AuDHD brain to work differently & causing myself unnecessary extra stress with multiple failed attempts - I will look at ways to work with my brain better.

One my favorite Windows 10/11 keyboard shortcuts remains Win-H for "voice typing" or voice-to-text recognition. Easy to access, super accurate & built-in the OS.

Here's the list of special phrases to say for punctuation & voice commands:

https://support.microsoft.com/en-us/windows/use-voice-typing-to-talk-instead-of-type-on-your-pc-fec94565-c4bd-329d-e59a-af033fa5689f
#windows10 #windows11 #voicetotext #voicecommand #speechtotext #voicetyping

@Linux_Is_Best gibt ja jetzt LLM s um das Interface zu verbessern.. muss halt nur noch lokal werden.

#Voice2Text #VoiceToText #Google #Apple #Amazon #Voice #iOS #Android

@globcoco @kallekn Whisper https://github.com/openai/whisper is FOSS but not the data it was trained with. It's part of OpenAI. And I have used this implementation https://github.com/Nikorasu/LiveWhisper to transcribe Finnish phone calls. It's a python app that runs on a terminal, so not super user friendly yet but you run it and it listen to your mic and transcribe what it hears and optionally translate it
#Whisper #LiveWhisper #OpenAI #Transcription #VoiceToText #Translate #Language

I'm just full of questions today. How can I easily make a transcript of a podcast without literally having to type up all the stuff?

What voice to text thing is really good?

#VoiceToText

#Security #Nerds and #Experts of #Mastodon, I have a #Question for you.
.
.
I am attempting to de-#Google my digital life (as much as is possible at the moment). I got myself an #Android #MobilePhone that does not have Google services installed on it, and I am using #FDroid, #AuroraStore, #APKPure, etc. to get the apps, and using the #AppBrain #AdDetector and #DuckDuckGo's #App #Tracking #Protection (#Beta built into their mobile browser).
.
I am doing my best to use #minimal apps, and only getting ones that are not security focused when necessary. I am not logging into Google, or any other #BigSocial/ #BigData services except when necessary.
.
My question is, are there any #SMS / #Texting apps that allow you to login on the #Web and #sync messages to your phone (even if only on the same network), and a #keyboard that has #VoiceToText that does not use Google or #Microsoft? I am currently using #SwiftKey, but only until I figure out the #T9 app I am using with the KB (HYBRID phone running Android 11).

Better #offline #opensource #SpeechRecognition using state-of-the-art whisper AI. Talk into mic and it types on screen! Translates all languages into English text.
https://github.com/themanyone/voice_typing

#VoiceTechnology #Transcription #DictationSoftware #VoiceToText #SpeechToText #DictationApp #VoiceRecognition #TranscriptionService #TextToSpeech #Translation #Dictation