A comprehensive computational model of animal biosonar signal processing

Abstract

Computational models of animal biosonar seek to identify critical aspects of echo processing responsible for the superior, real-time performance of echolocating bats and dolphins in target tracking and clutter rejection. The Spectrogram Correlation and Transformation (SCAT) model replicates aspects of biosonar imaging in both species by processing wideband biosonar sounds and echoes with auditory mechanisms identified from experiments with bats. The model acquires broadband biosonar broadcasts and echoes, represents them as time-frequency spectrograms using parallel bandpass filters, translates the filtered signals into ten parallel amplitude threshold levels, and then operates on the resulting time-of-occurrence values at each frequency to estimate overall echo range delay. It uses the structure of the echo spectrum by depicting it as a series of local frequency nulls arranged regularly along the frequency axis of the spectrograms after dechirping them relative to the broadcast. Computations take place entirely on the timing of threshold-crossing events for each echo relative to threshold-events for the broadcast. Threshold-crossing times take into account amplitude-latency trading, a physiological feature absent from conventional digital signal processing. Amplitude-latency trading transposes the profile of amplitudes across frequencies into a profile of time-registrations across frequencies. Target shape is extracted from the spacing of the object’s individual acoustic reflecting points, or glints, using the mutual interference pattern of peaks and nulls in the echo spectrum. These are merged with the overall range-delay estimate to produce a delay-based reconstruction of the object’s distance as well as its glints. Clutter echoes indiscriminately activate multiple parts in the null-detecting system, which then produces the equivalent glint-delay spacings in images, thus blurring the overall echo-delay estimates by adding spurious glint delays to the image. Blurring acts as an anticorrelation process that rejects clutter intrusion into perceptions.Author summaryBats and dolphins use their biological sonar as a versatile, high-resolution perceptual system that performs at levels desirable in man-made sonar or radar systems. To capture the superior real-time capabilities of biosonar so they can be imported into the design of new man-made systems, we developed a computer model of the sonar receiver used by echolocating bats and dolphins. Our intention was to discover the processing methods responsible for the animals’ ability to find and identify targets, guide locomotion, and prevent classic types of sonar or radar interference that hamper performance of man-made systems in complex, rapidly-changing surroundings. We have identified several features of the ears, hearing, time-frequency representation, and auditory processing that are critical for organizing echo-processing methods and display manifested in the animals’ perceptions.

Faster Repetition Rate Sharpens the Cortical Representation of Echo Streams in Echolocating Bats

“…Furthermore, the receptive fields of target-range (pulse-echo delay-tuned) neurons in fruit bat auditory cortex became sharper with increasing repetition rate (O'Neill and Suga, 1982;Wong et al, 1992;Tanaka and Wong, 1993). Recent computational models of animal biosonar proposed that the bat auditory cortex is likely to encode echo spectral details by transposing amplitude-latency trade-offs in the ascending auditory system into a topographical profile of spike time-registrations (Simmons, 2012;Ming et al, 2021). Here, we show that in insectivorous bats, the forward suppression induced by increasing pulse repetition rate sharpened the resolution of these time registrations, and thereby enhanced the cortical ensemble representation of echo spectral envelope, which may be particularly important for bats hunting insects on the wing.…”

“…Independently derived computational models also predict that bat auditory cortex relies upon a precise spike-time dependent synchronization network to reconstruct echoes and classify their source (Saillant et al, 1993;Ming et al, 2021). There is evidence that forward suppression contribute to a sharper rate coding of echo delay (Beetz et al, 2016).…”

“…Although there is extensive literature about the use of spike rate to encode temporal parameters, like duration or delay tuning, the bat auditory cortex also uses time coding strategies to cope with the rapid and temporally precise nature of echolocation (Suga, 1989). We took advantage of this to measure the effects of forward suppression on the cortical representation of complex sounds within the context of a computational model of information coding in auditory cortex (Ming et al, 2021). The bat auditory system is adapted to analyze biosonar echoes (Suga, 1989), which are typically 2-3 ms broadband, downward frequency-modulated sweeps.…”

“…When an echolocation pulse is reflected off of an irregularly shaped surface, multiple overlapping echoes convolve into a single sound endowed with a complex interference pattern of spectral peaks and notches (Simmons et al, 1974). Echo spectral notch patterns are therefore unique to each target, and bats can reconstruct an internal representation of a target's shape based on these fine spectral details (Schmidt, 1988;Simmons and Chen, 1989;Schmidt, 1992;Ming et al, 2021). The bat auditory cortex contains neurons that preferentially respond with greater spike rates to the presence of unique spectral interference patterns (Sanderson and Simmons, 2002;Firzlaff and Schuller, 2007).…”

Visualization of bat “echo space” using acoustic simulation

“…For instance, we should include the directivity of the ear position by introducing the head-related transfer function (HRTF) of the bat, including its ears, in the acoustic simulation space and a moving source to reflect the effect of the Doppler shift during flight to obtain more detailed spatial information. In addition, a bat auditory processing algorithm should be introduced into the post-echo processing such as the spectrogram correlation and transformation (SCAT) model [35][36][37] . We believe that this approach of echo simulation is a useful first step towards elucidating the perception space by bats.…”