Yicheng (@ChrisYicheng)

책 같은 정적 콘텐츠는 한 번만 음성을 생성해 CDN에 저장하면 요청당 비용을 사실상 없앨 수 있다는 제안과 함께 작성자가 로컬에서 학습한 TTS 모델이 Kokoro보다 성능이 더 좋다고 밝힘. 'listen' 기능 구현을 희망하며 구현에 도움을 제공하겠다고 함.

https://x.com/ChrisYicheng/status/2028359007854002659

#tts #texttospeech #audio #model #cdn

Yicheng (@ChrisYicheng) on X

@deedydas Cool! Would love to see a "listen" feature here, and the cost problem is solvable. Books are static content: generate audio once, store on CDN, serve forever. Zero per-request cost. We actually trained a local TTS model that outperforms Kokoro. Happy to help make this happen.

X (formerly Twitter)
Chatterbox Turbo is amazing! Long are gone the days I had to use Google #TTS / Speech Synthesis to listen to my ebooks. First it was Kokoro TTS, then F5-TTS but now Chatterbox Turbo is king. It sounds realistic enough, but the best thing is that it is Fast. I did an entire ebook in about an hour and a half.
Perfect for the #webNovels and #lightnovels I like to read.

@nclick sehr gut. Vielen Dank. Werde ich testen. Ich habe schon einige alternative #naviapps getestet. Mit den bekannten wie #OsmAnd #GraphHopper #organicmaps bin ich nicht wirklich happy gewesen und habe bisher #MagicEarth für mich entdeckt. #CoMaps runtergeladen und installiert! Danke f Hinweis!

PS: Die #Sprachausgabe #TTS ist in meinem E/OS shitty. (Gibt es eine nette alternative TTS-Plastikdame in deutsch?)

#navigation #naviapp #did #didit #diday #digitalindependenceDay #eos #android

hey guys, for all you #tts #nvda fokes, quick question. do any of you know a way to take NVDA speech and to turn it into downloadable mp3/wav?
I realize I'm biased, but tell me this isn't one of the coolest examples of local on device ML on Apple Silicon today? Very proud of this app: https://apps.apple.com/us/app/speaklone/id6758415075
#iOS #macoS #indiedev #mlx #tts #voice
There are also many AI TTS created by folks at 11 Labs and other open source implementations that support speech to text and text to speech. But I'm looking for traditional formant synthesis implementation. Do we have existing phonemizer or any sort of that that can aid Tamil TTS creation? #tamil #tts #linguist #language #formant (2/2)

Vincent van der Meulen (@vincentmvdm)

작성자는 PR(풀이 리퀘스트)을 일일이 읽기 귀찮아 Cursor AI의 클라우드 에이전트를 해킹해 데모 영상을 자동으로 내레이션하도록 만들었다고 보고합니다. 개인 음성 톤으로 TTS를 생성해 동료나 고객에게 보낼 수 있고, 필요한 것은 적절한 프롬프트뿐이라며 데모 자동화·음성 내레이션을 활용한 실전 적용 사례를 공유했습니다.

https://x.com/vincentmvdm/status/2027228328369774936

#cursor #cloudagents #tts #promptengineering

Vincent van der Meulen (@vincentmvdm) on X

i'm getting tired of reading prs... so i hacked @cursor_ai cloud agents to narrate the demos they send me. now all i have to do is watch a video :-) bonus: it's my own voice so i can send these to teammates or customers turns out all you need is a prompt! (see next tweet)

X (formerly Twitter)

Разрабатываем голосового ассистента на Rockchip. Часть 2

Продолжаю разрабатывать DIY голосового ассистента на SOC-платформе Rockchip. В первой части смы соединили в единый конвейер вызов распознавания речи, локального чат-бота и синтез ответа. Если еще не читали, то вам сюда . Во второй части поговорим об улучшениях работы с синтезом речи. Научим нашего ИИ-помощника произносить текст, содержащий сложные для моделей сущности, а также сделаем его более плавным.

https://habr.com/ru/companies/mts_ai/articles/1004144/

#ai #ml #voice #tts #voice_assistant #ииассистент #голосовой_помощник #голосовой_ассистент #голосовой_интерфейс #искусственный_интеллект

Разрабатываем голосового ассистента на Rockchip. Часть 2

Продолжаю разрабатывать DIY голосового ассистента на SOC-платформе Rockchip. В первой части мы соединили в единый конвейер вызов распознавания речи, локального чат-бота и синтез ответа. Если еще не...

Хабр

EyeingAI (@EyeingAI)

Noiz AI가 음성 클론을 3초 내 생성하고 감정 추가, 긴 텍스트 문장별 편집, 다국어 비디오 더빙을 몇 분 안에 처리할 수 있다고 발표했습니다. 게시자는 ElevenLabs가 압박받고 있다고 표현하며, 첫 달 프로모션으로 $1.9을 제시한다고 알립니다.

https://x.com/EyeingAI/status/2026649723604771176

#noizai #voicecloning #tts #dubbing

EyeingAI (@EyeingAI) on X

ElevenLabs is sweating lol Noiz AI now lets you clone a voice in 3 seconds, add emotion, edit long texts sentence by sentence, and dub videos in any language... all in mins. First month limited-time offer: $1.9 Let me show you with real demos: 👇

X (formerly Twitter)