Biomimetic detection of dynamic signatures in foliage echoes

Bhardwaj, Ananya; Khyam, Mohammad Omar; Müller, Rolf

doi:10.1088/1748-3190/abf910

Cited by 4 publications

(5 citation statements)

References 99 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since all the specific values of the parameters that the determine the reflection from an individual leaf remain unknown, a foliage is best approximated as a stochastic array of reflectors [27,29] that results in likewise unpredictable echo waveforms [27]. For sonar-based navigation, the implication of this randomness is that any sonar system will never see the same echo waveform again [30]. Hence, conventional template-based methods for recognition of a location-specific pattern will not work.…”

Section: Introductionmentioning

confidence: 99%

Small-scale location identification in natural environments with deep learning based on biomimetic sonar echoes

Zhang¹,

Farabow²,

Singhal³

et al. 2023

Bioinspir. Biomim.

Self Cite

View full text Add to dashboard Cite

Many bat species navigate in complex, heavily vegetated habitats. To achieve this, the animal relies on a sensory basis that is very different from what is typically done in engineered systems that are designed for outdoor navigation. Whereas the engineered systems rely on data-heavy senses such as lidar, bats make do with echoes triggered by short, ultrasonic pulses. Prior work has shown that "clutter echoes" originating from vegetation can convey information on the environment they were recorded in -- despite their unpredictable nature. The current work has investigated the spatial granularity that these clutter echoes can convey by applying deep-learning location identification to an echo data set that resulted from the dense spatial sampling of a forest environment. The GPS location corresponding to the echo collection events was clustered to break the survey area into the number of spatial patches ranging from two to 100. A convolutional neural network (Resnet 152) was used to identify the patch associated with echo sets ranging from one to ten echoes. The results demonstrate a spatial resolution that is comparable to the accuracy of recreation-grade GPS operating under foliage cover. This demonstrates that fine-grained location identification can be accomplished at very low data rates.

show abstract

Section: Introductionmentioning

confidence: 99%

Small-scale location identification in natural environments with deep learning based on biomimetic sonar echoes

Zhang¹,

Farabow²,

Singhal³

et al. 2023

Bioinspir. Biomim.

Self Cite

View full text Add to dashboard Cite

show abstract

“…This makes the waveforms of individual echoes exceedingly hard to predict. As a result of the unpredictable and irreproducible nature, a traditional correlation analysis of natural foliage echoes did not pick up any common patterns in echo waveforms recorded from forests beyond the pulse that elicited them [39]. These findings do not bode well for traditional pattern recognition methods that depend on deterministic templates and linear dependencies that can be picked up by correlation measures.…”

Section: Introductionmentioning

confidence: 81%

“…Correlation coefficients between the recorded echo waveforms were calculated to make sure the echoes contained invariant information during the site's data collection. A confusion matrix used to show the correlation between each pair of echoes along the foliage trial recordings, 50 continued recording echo examples correlation coefficient (mean and standard deviation) were calculated along one of the trail, about 10 m. As a reference for the experimental correlation data, simulated echoes were used to calculate the correlation coefficient for random independently distributed impulse responses that were convolved with the same pulse template that was used to generate the foliage echoes in the field experiments [39]. The waveforms of the simulated echoes had the same assumed sampling rate (400 kHz) and duration (10 ms) as the physical echo data.…”

Section: Signal Processingmentioning

confidence: 99%

Large-scale recognition of natural landmarks with deep learning based on biomimetic sonar echoes

Zhang¹,

Müller²

2022

Bioinspir. Biomim.

View full text Add to dashboard Cite

The ability to identify natural landmarks on a regional scale could contribute to the navigation skills of echolocating bats and also advance the quest for autonomy in natural environments with man-made systems. However, recognizing natural landmarks based on biosonar echoes has to deal with the unpredictable nature of echoes that are typically superpositions of contributions from many different reﬂectors with unknown properties. The results presented here show that a deep neural network (ResNet50) was able to classify 10 different ﬁeld sites and 20 different tracks (2 at each site) distributed over an area about 40 kilometers in diameter. Based on spectrogram representations of single echoes, classiﬁcation accuracies up to 99.6% for different sites and 94.7% for different tracks have been achieved. Classiﬁcation performance was found to depend on the used pulse component (constant-frequency - CF vs. frequency-modulated - FM) and the trade-off between time and frequency resolution in the spectrogram representations of the echoes. For the former, classiﬁcation performance increased monotonically with better time resolution. For the latter, classiﬁcation performance peaked at an intermediate trade-off point between time and frequency resolution indicating that both dimensions contained relevant information. Future work will be needed to further characterize the quality of the spatial information contained in the echoes, e.g., in terms of spatial resolution and potential ambiguities.

show abstract

“…Aside from challenges with practical applications of audio transducers, further research is required to better characterize vegetation echoes. Plant echoes have mainly been studied in the context of bat acoustics (Danilovich et al, 2020; Kohles et al, 2020) and bio‐sonar development (Bhardwaj et al, 2021; Zhang & Müller, 2022), but further research is needed to examine ways to convert plant echoes into signal proxies of overall plant community health.…”

Section: Challenges and Constraintsmentioning

confidence: 99%

Application of active acoustic transducers in monitoring and assessment of terrestrial ecosystem health—A review

Rostami

Nansen

2022

Methods Ecol Evol

View full text Add to dashboard Cite

1. Urbanization, agricultural production, natural resource extractions and climate change are global drivers of terrestrial ecosystem degradation and decline in ecosystem health. Assessing vegetation structures provides valuable direct and indirect information on biodiversity and ecosystem health, since plant communities govern terrestrial ecosystem functions and overall biodiversity. 2. A growing body of literature supports the hypothesis that acoustic profiling and characterization of soundscapes may reveal natural patterns and ecosystem responses to environmental disturbances. Although assessments of ecosystem soundscapes are a promising tool for ecological monitoring, the influence of vegetation structure on acoustic indices remains largely unaddressed. An effective non-invasive approach to monitoring ecosystem health includes use of active acoustic transducers, which convert mechanical/acoustic energy into electrical energy and visa-versa.3. This article reviews and discusses possible applications (and also constraints) of active acoustic transducers in monitoring and assessments of terrestrial ecosystem health. Specifically, this article includes a brief introduction to the basic principles of sound and types of active acoustic transducers. Moreover, we provide reviews of common uses of active acoustic transducers in assessing plant structures and plant functional traits.4. We emphasize that active acoustic transducers can be used to analyse plant characteristics in a remote, scalable, non-invasive and cost-effective manner to characterize overall terrestrial ecosystem health. Suggestions and recommendations for future research and management directions that may facilitate applications of acoustic transducers to natural terrestrial landscapes are provided.

show abstract

Biomimetic detection of dynamic signatures in foliage echoes

Cited by 4 publications

References 99 publications

Small-scale location identification in natural environments with deep learning based on biomimetic sonar echoes

Small-scale location identification in natural environments with deep learning based on biomimetic sonar echoes

Large-scale recognition of natural landmarks with deep learning based on biomimetic sonar echoes

Application of active acoustic transducers in monitoring and assessment of terrestrial ecosystem health—A review

Contact Info

Product

Resources

About