Hume AI has introduced a new feature called Voice Control, designed to provide developers with precise control over AI voice customization. This experimental tool allows for the creation of unique AI voices by adjusting base voices along ten interpretable dimensions, such as nasality, gender expression, and buoyancy. Unlike traditional prompt-based methods, Voice Control offers continuous adjustments, enabling reproducible voice modifications across sessions. This feature is part of Hume's Empathic Voice Interface (EVI) and is currently available in beta.
Introducing Voice Control by Hume
— Hume (@hume_ai) December 2, 2024
We developed an experimental voice modulation approach that enables you to create unique AI voices in seconds.
Our voice sliders make it intuitive to adjust base voices along 10 interpretable dimensions including:
👃 Nasality: resonant to… pic.twitter.com/zqWy05OrlN
The release of Voice Control addresses the challenge of finding suitable AI voices for brands and applications without compromising quality or resorting to risky voice cloning techniques. By using an intuitive no-code interface, developers can craft personalized voices that align with their brand's identity. The tool ensures that changes to one voice characteristic do not affect others, maintaining the integrity of the base voice.
Hume AI, a startup focused on integrating emotional intelligence into AI systems, is behind this development. The company aims to enhance human-computer interaction by making AI more empathetic and responsive to emotional cues. This aligns with their broader vision of optimizing AI for human well-being.
