Amphion is an open-source toolkit designed to support research and development in audio, music and speech generation. According to the project’s GitHub site, it offers unique visualizations of classic models and architectures to help junior researchers and engineers better understand them. The toolkit supports various individual generation tasks such as text-to-speech (TTS), singing voice synthesis …
Useful, an AI company focused on improving human-machine communication, has open-sourced Moonshine, a new speech-to-text model that aims to significantly reduce the latency of voice interfaces. According to Useful founder Pete Warden, Moonshine returns results 1.7 times faster than OpenAI’s Whisper model while matching or exceeding its accuracy. The model’s variable-length input window allows it …
Amazon has introduced a generative AI tool that enables brands to create audio ads in addition to images and videos, expanding its advertising offerings at the Amazon unBoxed conference. As AdWeek reports, this new feature allows advertisers to generate ads using minimal product information and is part of a broader suite of tools aimed at …
Gladia, an AI-powered transcription and audio intelligence provider, has secured $16 million in funding. The Paris-based company plans to develop a new real-time transcription and analytics engine with this investment. CEO Jean-Louis Quéguiner told VentureBeat he founded the company out of frustration with existing services’ poor accent recognition. Gladia’s new engine can transcribe over 100 …
Play.ht has released its new speech model, “Play 3.0 mini.” This AI-powered text-to-speech model can speak in over 30 languages and imitate any voice or accent. It offers industry-leading speed and accuracy, according to Play.ht. Play 3.0 mini was designed specifically for conversational AI applications and is intended to be particularly reliable and cost-efficient. The …
AI startup Rep.ai has raised $7.5 million in funding to launch its “digital twin” technology for sales representatives. This was reported by Michael Nuñez for VentureBeat. The company, formerly known as ServiceBell, is developing AI-powered avatars to assist website visitors in real-time video and audio conversations. Rep.ai combines visual and vocal replication with natural language …
Researchers at Johns Hopkins University and Tencent AI Lab have developed a new text-to-audio model called EzAudio. As Michael Nuñez reports for VentureBeat, EzAudio can generate high-quality sound effects from text descriptions. The model uses an innovative method for processing audio data and a new architecture called EzAudio-DiT. In tests, EzAudio outperformed existing open-source models …
Hume AI has unveiled an improved version of its voice AI EVI 2. The new version offers more natural conversations, faster response times, and more voice customization options. This was reported by Carl Franzen for VentureBeat. According to Hume co-founder Alan Cowen, EVI 2 can now be integrated directly into apps to process user requests. …
Google is adding an audio feature to its AI-powered note-taking app NotebookLM. The new “Audio Overview” allows users to get verbal explanations of complex topics from uploaded documents, as Aisha Malik reports for TechCrunch. AI-generated virtual hosts summarize the content and explain difficult concepts using metaphors. The feature is aimed at people who absorb information …
The music AI Suno is now available for iPhones in the USA. It offers a variety of styles and genres, and can create full songs with lyrics and vocals, as well as instrumentals. The app will soon be available in other countries and for Android devices.