Generic HRTFs May be Good Enough in Virtual Reality. Improving Source Localization through Cross-Modal Plasticity

Area of Science:

• Human perception
• Virtual Reality (VR) technology
• Acoustic spatialization

Background:

• Auditory spatial localization relies on interaural time/level differences and spectral cues.
• Virtual reality (VR) uses Head Related Transfer Functions (HRTFs) for spatialized sound, with individualized HRTFs offering accuracy but requiring calibration, while generic HRTFs are convenient but less precise.
• Human auditory space representation is known to be plastic.

Purpose of the Study:

• To investigate if cross-modal learning can enhance auditory source localization using generic HRTFs in VR.
• To determine the effectiveness of synchronized audiovisual stimuli in improving spatial hearing within VR environments.

Main Methods:

• Participants used generic HRTFs for auditory spatial localization tasks.
• Experimental groups were exposed to either dynamic auditory stimuli with aligned visual counterparts, auditory stimuli alone, or asynchronous audiovisual stimuli.
• Post-exposure auditory source localization performance was assessed.

Main Results:

• Pairing dynamic auditory stimuli with spatio-temporally aligned visual cues significantly improved auditory source localization for users of generic HRTFs.
• Exposure to auditory stimuli alone or asynchronous audiovisual stimuli did not yield improvements in auditory localization.
• Cross-modal learning via synchronized audiovisual input enhances the efficacy of generic HRTFs.

Conclusions:

• Auditory spatial localization in VR can be improved using generic HRTFs through cross-modal learning with synchronized audiovisual stimuli.
• These findings suggest that generic HRTFs may be sufficient for achieving good auditory localization in VR, reducing the need for complex individual calibrations.
• The plasticity of acoustic space representation is further supported, with implications for VR system development and human perception research.