التقنية والذكاء الاصطناعي
Voice AI in vogue: Nvidia releases new open transcription model
المصدر: TechCrunch · نُشر
Nvidia just released a free speech-to-text model that can transcribe up to 40 languages, but it only works on recordings, not live audio, so it’s a big step for meeting and medical transcription tools while still leaving real-time use to rivals like ElevenLabs.
Nvidia just released a free speech-to-text model that can transcribe up to 40 languages, but it only works on recordings, not live audio, so it’s a big step for meeting and medical transcription tools while still leaving real-time use to rivals like ElevenLabs.
أخبار ذات صلة
DeepMind unveils Genie 4 world model promising more realistic physics and humans
Last year, Google DeepMind sent gaming stocks tumbling with the reveal of its Genie 3 world model . Since then, world models have become a major focus of AI labs around the world, but Google isn't resting on its laurels. Researchers have been hard at work on an updated model that
Ars Technica - All content ·
Reddit is the latest platform to embrace the ‘For You’ feed
The For You feed combines posts from communities users have joined with recommendations tailored to their interests.
TechCrunch ·
Reddit is getting a For You feed
Reddit is switching up its homepage feeds. The company announced today that it's going to start offering a "new feed experience" that includes both a Following feed to show you posts from subreddits you follow and a For You feed that includes posts from your subreddits and "recom
The Verge ·
Reddit is adding a 'for you' feed for recommendations
The app will have a "following" tab to surface content from subreddits you've joined.
Engadget - Technology News & Expert Reviews ·
AI execs may be too out of touch to build for the general public
Tech leaders are flaunting how out of touch they are when they brag about using AI to connect with their families.
TechCrunch ·