Object and setting identification in natural auditory scenesa)

Area of Science:

• Auditory perception
• Cognitive psychology
• Acoustic scene analysis

Background:

• Auditory scene perception enables identification of environmental settings and objects, crucial for situational awareness.
• While visual scene recognition is well-researched, auditory scene and object identification in complex environments remain less understood.

Purpose of the Study:

• To investigate the impact of auditory scene duration on the identification of environmental settings and specific objects.
• To compare the accuracy of setting versus object identification as a function of scene duration.

Main Methods:

• Participants listened to natural auditory scenes presented at varying durations (1, 2, and 4 seconds).
• Listeners were tasked with identifying the environmental setting (e.g., café) and discrete objects (e.g., music, talking) within each scene.
• Statistical models incorporating low- and mid-level acoustic features were used to predict performance.

Main Results:

• Object identification accuracy was consistently higher than setting identification accuracy across all tested durations.
• Increased scene duration significantly improved performance for both setting and object identification tasks.
• Acoustic features predicted performance but explained a limited portion of the overall variance.

Conclusions:

• Auditory scene duration is a critical factor influencing both setting and object identification.
• Differences in performance suggest that setting identification relies more on integrating global scene properties and multiple sound sources, while object identification emphasizes individual sound source segregation and feature recognition.
• Perceptual and cognitive processes interact significantly in auditory scene and object recognition.