Discrimination of object information by bat echolocation deciphered from acoustic simulations

Area of Science:

• Bioacoustics
• Sensory Ecology
• Animal Behavior

Background:

• High-precision visual sensing with deep learning faces limitations in detecting occluded information.
• Optical sensors struggle with objects hidden from view, a challenge not faced by longer wavelength sound.
• Bats utilize echolocation, a sophisticated form of sound sensing, for navigation and foraging.

Purpose of the Study:

• To investigate if bats (Pipistrellus abramus) can acquire occlusion information using echolocation.
• To analyze the acoustic cues bats use to differentiate targets based on hidden backend structures.

Main Methods:

• A two-alternative forced choice test was performed on bats using targets with identical fronts but varied back geometries.
• Three-dimensional acoustic simulations computed echo impulse responses for detailed analysis of acoustic cues.
• Analysis focused on occlusion-related components within simulated echo impulse responses.

Main Results:

• Bats successfully discriminated between targets based on differences in occlusion spot structure and texture via echolocation.
• Discrimination performance correlated with spectral distortion in occlusion-related echo components.
• Simulated echo impulse responses provided insights into acoustic cues used by bats.

Conclusions:

• Bats effectively utilize echolocation to perceive occluded information, including geometric and textural differences.
• Echolocation provides bats with access to target details beyond the reach of optical sensing.
• Broadband ultrasound sensing offers significant advantages for detailed environmental information acquisition.