Browsing by Subject "Acoustic"
- Results Per Page
- Sort Options
Item Open Access An explainable COVID-19 detection system based on human sounds.(Smart health (Amsterdam, Netherlands), 2022-12) Li, Huining; Chen, Xingyu; Qian, Xiaoye; Chen, Huan; Li, Zhengxiong; Bhattacharjee, Soumyadeep; Zhang, Hanbin; Huang, Ming-Chun; Xu, WenyaoAcoustic signals generated by the human body have often been used as biomarkers to diagnose and monitor diseases. As the pathogenesis of COVID-19 indicates impairments in the respiratory system, digital acoustic biomarkers of COVID-19 are under investigation. In this paper, we explore an accurate and explainable COVID-19 diagnosis approach based on human speech, cough, and breath data using the power of machine learning. We first analyze our design space considerations from the data aspect and model aspect. Then, we perform data augmentation, Mel-spectrogram transformation, and develop a deep residual architecture-based model for prediction. Experimental results show that our system outperforms the baseline, with the ROC-AUC result increased by 5.47%. Finally, we perform an interpretation analysis based on the visualization of the activation map to further validate the model.Item Open Access Bayesian Techniques for Adaptive Acoustic Surveillance(2010) Morton, Kenneth DAutomated acoustic sensing systems are required to detect, classify and localize acoustic signals in real-time. Despite the fact that humans are capable of performing acoustic sensing tasks with ease in a variety of situations, the performance of current automated acoustic sensing algorithms is limited by seemingly benign changes in environmental or operating conditions. In this work, a framework for acoustic surveillance that is capable of accounting for changing environmental and operational conditions, is developed and analyzed. The algorithms employed in this work utilize non-stationary and nonparametric Bayesian inference techniques to allow the resulting framework to adapt to varying background signals and allow the system to characterize new signals of interest when additional information is available. The performance of each of the two stages of the framework is compared to existing techniques and superior performance of the proposed methodology is demonstrated. The algorithms developed operate on the time-domain acoustic signals in a nonparametric manner, thus enabling them to operate on other types of time-series data without the need to perform application specific tuning. This is demonstrated in this work as the developed models are successfully applied, without alteration, to landmine signatures resulting from ground penetrating radar data. The nonparametric statistical models developed in this work for the characterization of acoustic signals may ultimately be useful not only in acoustic surveillance but also other topics within acoustic sensing.
Item Open Access Quantifying vocal response in experimental playbacks to Risso's dolphins(2014-04-23) Boucher, AimeeIn a world of constant technological development and expansion into the marine environment, the marine soundscape is constantly changing. With the addition of anthropogenic sources from naval sonar to seismic survey vessels over the past century, the deficiency of knowledge on the impact of such acoustic disturbance leaves little guidance for effective regulation of anthropogenic marine noise pollution. To help address this, the U.S. Department of Defense’s Strategic Environmental Research and Development Program (SERDP) has teamed with multiple academic and scientific institutions to research and catalog the baseline behavioral ecology across a range of odontocete species, which can then serve as a baseline for additional research. This report examines a portion of that project, conducted to assess the response of Risso’s dolphin, Grampus griseus, to natural stimuli. During an August 2013 playback study off Southern California, acoustic data were collected via digital acoustic recording tags (DTAGs) to identify the vocal response of three Risso’s dolphins, Grampus griseus. The playbacks consisted of calls from three cetacean species: Megaptera novaeangliae, Orcinus orca, and Grampus griseus. To determine whether the vocal rate measurements could be reliably quantified, a repeatability experiment was conducted. Two playback studies (O. orca and G. griseus) were conducted on one animal, while three playbacks (O. orca, G. griseus, and M. novaeangliae) were presented to two animals. Only one of the tagged animals demonstrated a noteworthy response to the O. orca exposure, with more than a 500% increase in vocalizations after the playback. Vocal rate did not vary considerably in the tag with O. orca and G. griseus playbacks and the other tag resulted in roughly zero vocalizations during pre- and post-playbacks. Based on the small sample size, it appears that G. griseus response varies in the presence of a predator – with one tag demonstrating a dramatic increase of vocal rate when exposed to O. orca calls. These results are a necessary early step in gathering baseline information on the behavioral ecology of cetaceans susceptible to anthropogenic acoustic impact. A continuation of this project and further research is necessary to fully understand how marine mammals perceive and are impacted by human expansion into the marine soundscape.