Visual mechanisms for voice-identity recognition flexibly adjust to auditory noise level

Area of Science:

• Neuroscience
• Auditory Perception
• Visual Cognition

Background:

• Voice identity recognition is crucial for communication.
• A 'face-benefit' enhances voice recognition when voices are learned with faces.
• This benefit is amplified in noisy conditions, but neural mechanisms remain unclear.

Purpose of the Study:

• To investigate the neural basis of the face-benefit in voice recognition under varying auditory noise levels.
• To examine the role of face-sensitive brain regions in auditory-only voice recognition.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to study brain activity.
• Participants recognized auditory-only speakers previously learned by face.
• Experiments were conducted in high (-4 dB SNR) and low (+4 dB SNR) noise conditions.

Main Results:

• A behavioral face-benefit was observed in both noise levels.
• In high noise, the right posterior superior temporal sulcus motion-sensitive face area (pSTS-mFA) was engaged, with increased connectivity to voice areas.
• In low noise, the fusiform face area (FFA) and right pSTS-mFA showed increased responses.

Conclusions:

• The visual network adapts to support voice-identity recognition even without visual input.
• The pSTS-mFA plays a key role in high-noise voice recognition, potentially integrating visual cues.
• The FFA is involved in voice recognition, particularly in less noisy environments.