_Audio AI Model Company Joins the Metaverse, Noiz Completes Multi-Million Dollar Seed Round_

According to Dongcha Beating monitoring, the audio AI company Noiz has completed tens of millions of dollars in seed funding, with the latest investor being GGV Capital. Unlike most audio models, it is not planning to continue on music generation, speech recognition, or TTS. Instead, it aims to enable AI to understand the physical world behind sound.

For example, when hearing a door closing, the model should not only recognize "the door is closed" but also understand what material the door is, how much force was used, where the sound is coming from, and how the room's structure affects the sound. Noiz is specifically collecting data containing spatial, material, distance, and directional information.

Its ultimate goal is to create more of a "soundscape world model": given the current environment and events taking place, the model can deduce what should be heard at that moment. This will allow audio models to go from dubbing and sound effect tools to further integrating real-time games, virtual worlds, and robots, enabling machines to understand what is happening around them through sound.