Text-to-speech model

Smallest.ai, a San Francisco, California-based artificial intelligence startup focused on multimodal modeling, has introduced Lightning text-to-speech (TTS) models capable of generating up to 10 seconds of audio in less than 100 milliseconds, PR Newswire reports. This advancement enables developers around the world to build highly realistic speech robotics applications with sub-second latency, streamlining the implementation process and making it accessible at a significantly lower cost. Lightning currently supports multiple accents of English and Indian...

A team of Amazon AI researchers has announced the development of what is claimed to be the largest text-to-speech model ever, with the most parameters and using the largest training dataset. The researchers have published a paper on the arXiv preprint server detailing the development and training process of the model. In recent years, "large-scale language models" like ChatGPT have gained attention for their ability to intelligently answer questions and generate high-level text. However, AI is also being integrated into other mainstream application areas. In this new project, researchers try to increase the number of parameters and expand...