Blind Spatial Impulse Response Generation from Separate Room- and Scene-Specific Information
This addresses the challenge of realistic audio rendering in AR applications where direct acoustic measurements are impractical, offering a solution for seamless sound integration in dynamic environments.
The paper tackles the problem of generating spatial room impulse responses for augmented reality audio without acoustic measurements by inferring room-specific information from available sound sources and using it to render new sources at different positions. The result is a method that combines an encoder network with a diffusion-based generator to produce responses that incorporate both room- and position-specific parameters.
For audio in augmented reality (AR), knowledge of the users' real acoustic environment is crucial for rendering virtual sounds that seamlessly blend into the environment. As acoustic measurements are usually not feasible in practical AR applications, information about the room needs to be inferred from available sound sources. Then, additional sound sources can be rendered with the same room acoustic qualities. Crucially, these are placed at different positions than the sources available for estimation. Here, we propose to use an encoder network trained using a contrastive loss that maps input sounds to a low-dimensional feature space representing only room-specific information. Then, a diffusion-based spatial room impulse response generator is trained to take the latent space and generate a new response, given a new source-receiver position. We show how both room- and position-specific parameters are considered in the final output.