An official website of the United States government
Flag flutter frequently features a marked difference between the onset speed of flutter and the speed below which flutter stops. The hysteresis tends to be especially large in experiments as opposed to simulations. This phenomenon has been ascribed to inherent imperfections of flatness in experimental samples, which are thought to inhibit the onset of flutter but have a lesser effect once a flag is already fluttering. In this work, we present an experimental confirmation for this explanation through motion tracking. We also visualize the wake to assess the potential contribution of discrete vortex shedding to hysteresis. We then mould our understanding of the mechanism of bistability and additional observations on flag flutter into a novel, observation-based, semiempirical model for flag flutter in the form of a single ordinary differential equation. Despite its simplicity, the model successfully reproduces key features of the physical system such as bistability, sudden transitions between non-fluttering and fluttering states, amplitude growth and frequency growth.
more »
« less
ADAPTIVE TIME–FREQUENCY SCATTERING FOR PERIODIC MODULATION RECOGNITION IN MUSIC SIGNALS
Vibratos, tremolos, trills, and flutter-tongue are techniques frequently found in vocal and instrumental music. A com- mon feature of these techniques is the periodic modulation in the time–frequency domain. We propose a representa- tion based on time–frequency scattering to model the inter- class variability for fine discrimination of these periodic modulations. Time–frequency scattering is an instance of the scattering transform, an approach for building invari- ant, stable, and informative signal representations. The proposed representation is calculated around the wavelet subband of maximal acoustic energy, rather than over all the wavelet bands. To demonstrate the feasibility of this approach, we build a system that computes the represen- tation as input to a machine learning classifier. Whereas previously published datasets for playing technique analy- sis focus primarily on techniques recorded in isolation, for ecological validity, we create a new dataset to evaluate the system. The dataset, named CBF-periDB, contains full- length expert performances on the Chinese bamboo flute that have been thoroughly annotated by the players them- selves. We report F-measures of 99% for flutter-tongue, 82% for trill, 69% for vibrato, and 51% for tremolo detec- tion, and provide explanatory visualisations of scattering coefficients for each of these techniques.
more »
« less
- Award ID(s):
- 1633206
- PAR ID:
- 10118936
- Date Published:
- Journal Name:
- International Society for Music Information Retrieval Conference
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Convolutional neural networks (CNNs) are revolutionizing imaging science for two- and three-dimensional images over Euclidean domains. However, many data sets are intrinsically non-Euclidean and are better modeled through other mathematical structures, such as graphs or manifolds. This state of affairs has led to the development of geometric deep learning, which refers to a body of research that aims to translate the principles of CNNs to these non-Euclidean structures. In the process, various challenges have arisen, including how to define such geometric networks, how to compute and train them efficiently, and what are their mathematical properties. In this letter we describe the geometric wavelet scattering transform, which is a type of geometric CNN for graphs and manifolds consisting of alternating multiscale geometric wavelet transforms and nonlinear activation functions. As the name suggests, the geometric wavelet scattering transform is an adaptation of the Euclidean wavelet scattering transform, first introduced by S. Mallat, to graph and manifold data. Like its Euclidean counterpart, the geometric wavelet scattering transform has several desirable properties. In the manifold setting these properties include isometric invariance up to a user specified scale and stability to small diffeomorphisms. Numerical results on manifold and graph data sets, including graph and manifold classification tasks as well as others, illustrate the practical utility of the approach.more » « less
-
We introduce a new theoretical approach for analyzing pump and probe experiments in non-linear systems of optical phonons. In our approach, the effect of coherently pumped polaritons is modeled as providing time-periodic modulation of the system parameters. Within this framework, propagation of the probe pulse is described by the Floquet version of Maxwell’s equations and leads to phenomena such as frequency mixing and resonant parametric production of polariton pairs. We analyze light reflection from a slab of insulating material with a strongly excited phonon-polariton mode and obtain analytic expressions for the frequency-dependent reflection coefficient for the probe pulse. Our results are in agreement with recent experiments by Cartella et al. [Proc. Natl. Acad. Sci. U. S. A. 115, 12148 (2018)], which demonstrated light amplification in a resonantly excited SiC insulator. We show that, beyond a critical pumping strength, such systems should exhibit Floquet parametric instability, which corresponds to resonant scattering of pump polaritons into pairs of finite momentum polaritons. We find that the parametric instability should be achievable in SiC using current experimental techniques and discuss its signatures, including the non-analytic frequency dependence of the reflection coefficient and the probe pulse afterglow. We discuss possible applications of the parametric instability phenomenon and suggest that similar types of instabilities can be present in other photoexcited non-linear systems.more » « less
-
ABSTRACT Variable active galactic nuclei showing periodic light curves have been proposed as massive black hole binary (MBHB) candidates. In such scenarios, the periodicity can be due to relativistic Doppler-boosting of the emitted light. This hypothesis can be tested through the timing of scattered polarized light. Following the results of polarization studies in type I nuclei and of dynamical studies of MBHBs with circumbinary discs, we assume a coplanar equatorial scattering ring, whose elements contribute differently to the total polarized flux, due to different scattering angles, levels of Doppler boost, and line-of-sight time delays. We find that in the presence of an MBHB, both the degree of polarization and the polarization position angle have periodic modulations. The polarization angle oscillates around the semiminor axis of the projected MBHB orbital ellipse, with a frequency equal either to the binary’s orbital frequency (for large scattering screen radii), or twice this value (for smaller scattering structures). These distinctive features can be used to probe the nature of periodic MBHB candidates and to compile catalogues of the most promising sub-pc MBHBs. The identification of such polarization features in gravitational-wave (GW) detected MBHBs would enormously increase the amount of physical information about the sources, allowing the measurement of the individual masses of the binary components, and the orientation of the line of nodes on the sky, even for monochromatic GW signals.more » « less
-
Neurological disabilities cause diverse health and mental challenges, impacting quality of life and imposing financial burdens on both the individuals diagnosed with these conditions and their caregivers. Abnormal brain activity, stemming from malfunctions in the human nervous system, characterizes neurological disorders. Therefore, the early identification of these abnormalities is crucial for devising suitable treatments and interventions aimed at promoting and sustaining quality of life. Electroencephalogram (EEG), a non-invasive method for monitoring brain activity, is frequently employed to detect abnormal brain activity in neurological and mental disorders. This study introduces an approach that extends the understanding and identification of neurological disabilities by integrating feature extraction, machine learning, and visual analysis based on EEG signals collected from individuals with neurological and mental disorders. The classification performance of four feature approaches—EEG frequency band, raw data, power spectral density, and wavelet transform—is assessed using machine learning techniques to evaluate their capability to differentiate neurological disabilities in short EEG segmentations (one second and two seconds). In detail, the classification analysis is conducted under two conditions: single-channel-based classification and region-based classification. While a clear demarcation between normal (healthy) and abnormal (neurological disabilities) EEG metrics may not be evident, their similarities and distinctions are observed through visualization, employing wavelet features. Notably, the frontal brain region (frontal lobe) emerges as a crucial area for distinguishing abnormalities among different brain regions. Also, the integration of wavelet features and visual analysis proves effective in identifying and understanding neurological disabilities.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
