Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model
Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, Artsiom Sanakoyeu
The Journal of the Audio Engineering Society (AES)
Head-related transfer-functions (HRTF) are a central part of spatializing audio. However measuring the near-field HRTF at close source distances presents unique challenges. In particular the existing sound sources designed to be appropriate for near-field HRTF measurements on human subjects exhibit a notable issue of being unable to generate a sufficient acoustic output level at lower frequencies (below 300 Hz) while keeping a proper omnidirectional directivity pattern at higher frequencies. This paper proposes a novel design to overcome this limitation of low-frequency range. Several aspects of the design were considered in the paper: type of enclosure, low-frequency extension, choice of transducers, and metrics for sound source assessment. The chosen solutions are discussed together with numerical and experimental verification. The source constructed under the design method and process described herein achieved a frequency range of 120–16,000 Hz for which it can be used to measure HRTFs at source distances as small as 0.15 m from the subject’s head.
Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, Artsiom Sanakoyeu
Bilge Acun, Benjamin Lee, Fiodar Kazhamiaka, Kiwan Maeng, Manoj Chakkaravarthy, Udit Gupta, David Brooks, Carole-Jean Wu
Ilkan Esiyok, Pascal Berrang, Katriel Cohn-Gordon, Robert Künnemann