Showing 1–15 of 15 results for author: Szorkovszky, A

Modeling Eye Gaze Velocity Trajectories using GANs with Spectral Loss for Enhanced Fidelity

Authors: Shailendra Bhandari, Pedro Lencastre, Rujeena Mathema, Alexander Szorkovszky, Anis Yazidi, Pedro Lind

Abstract: Accurate modeling of eye gaze dynamics is essential for advancement in human-computer interaction, neurological diagnostics, and cognitive research. Traditional generative models like Markov models often fail to capture the complex temporal dependencies and distributional nuance inherent in eye gaze trajectories data. This study introduces a GAN framework employing LSTM and CNN generators and disc… ▽ More Accurate modeling of eye gaze dynamics is essential for advancement in human-computer interaction, neurological diagnostics, and cognitive research. Traditional generative models like Markov models often fail to capture the complex temporal dependencies and distributional nuance inherent in eye gaze trajectories data. This study introduces a GAN framework employing LSTM and CNN generators and discriminators to generate high-fidelity synthetic eye gaze velocity trajectories. We conducted a comprehensive evaluation of four GAN architectures: CNN-CNN, LSTM-CNN, CNN-LSTM, and LSTM-LSTM trained under two conditions: using only adversarial loss and using a weighted combination of adversarial and spectral losses. Our findings reveal that the LSTM-CNN architecture trained with this new loss function exhibits the closest alignment to the real data distribution, effectively capturing both the distribution tails and the intricate temporal dependencies. The inclusion of spectral regularization significantly enhances the GANs ability to replicate the spectral characteristics of eye gaze movements, leading to a more stable learning process and improved data fidelity. Comparative analysis with an HMM optimized to four hidden states further highlights the advantages of the LSTM-CNN GAN. Statistical metrics show that the HMM-generated data significantly diverges from the real data in terms of mean, standard deviation, skewness, and kurtosis. In contrast, the LSTM-CNN model closely matches the real data across these statistics, affirming its capacity to model the complexity of eye gaze dynamics effectively. These results position the spectrally regularized LSTM-CNN GAN as a robust tool for generating synthetic eye gaze velocity data with high fidelity. △ Less

Submitted 5 December, 2024; originally announced December 2024.

Comments: 16

ACM Class: I.2; I.6

Journal ref: Nature Scientific reports 2025

Using neuronal models to capture burst and glide motion and leadership in fish

Authors: Linnéa Gyllingberg, Alex Szorkovszky, David J. T. Sumpter

Abstract: While mathematical models, in particular self-propelled particle (SPP) models, capture many of the observed properties of large fish schools, they do not always capture the interactions of smaller shoals. Nor do these models tend to account for the observation that, when swimming alone or in smaller groups, many species of fish use intermittent locomotion, often referred to as burst and coast or b… ▽ More While mathematical models, in particular self-propelled particle (SPP) models, capture many of the observed properties of large fish schools, they do not always capture the interactions of smaller shoals. Nor do these models tend to account for the observation that, when swimming alone or in smaller groups, many species of fish use intermittent locomotion, often referred to as burst and coast or burst and glide. Recent empirical studies have suggested that burst and glide movement is indeed pivotal to the social interactions of individual fish. In this paper, we propose a model of social burst and glide motion by combining a well-studied model of neuronal dynamics, the FitzHugh-Nagumo model, with a model of fish motion. We begin by showing that the model can capture the motion of a single fish swimming down a channel. By then extending to a two fish model, where visual stimuli of the position of the other fish affect the internal burst or glide state of the fish, we find that our model captures a rich set of swimming dynamics found in many species of fish. These include: leader-follower behaviour; periodic changes in leadership; apparently random (i.e. chaotic) leadership change; and pendulum-like tit-for-tat turn taking. Unlike SPP models, which assume that fish move at a constant speed, the model produces realistic motion of individual fish. Moreover, unlike previous studies where a random component is used for leadership switching to occur, we show that leadership switching, both periodic and chaotic, can be the result of a deterministic interaction. We give several empirically testable predictions on how fish interact and discuss our results in light of recently established correlations between fish locomotion and brain activity. △ Less

Submitted 3 April, 2023; originally announced April 2023.

Central pattern generators evolved for real-time adaptation to rhythmic stimuli

Authors: Alex Szorkovszky, Frank Veenstra, Kyrre Glette

Abstract: For a robot to be both autonomous and collaborative requires the ability to adapt its movement to a variety of external stimuli, whether these come from humans or other robots. Typically, legged robots have oscillation periods explicitly defined as a control parameter, limiting the adaptability of walking gaits. Here we demonstrate a virtual quadruped robot employing a bio-inspired central pattern… ▽ More For a robot to be both autonomous and collaborative requires the ability to adapt its movement to a variety of external stimuli, whether these come from humans or other robots. Typically, legged robots have oscillation periods explicitly defined as a control parameter, limiting the adaptability of walking gaits. Here we demonstrate a virtual quadruped robot employing a bio-inspired central pattern generator (CPG) that can spontaneously synchronize its movement to a range of rhythmic stimuli. Multi-objective evolutionary algorithms were used to optimize the variation of movement speed and direction as a function of the brain stem drive and the center of mass control respectively. This was followed by optimization of an additional layer of neurons that filters fluctuating inputs. As a result, a range of CPGs were able to adjust their gait pattern and/or frequency to match the input period. We show how this can be used to facilitate coordinated movement despite differences in morphology, as well as to learn new movement patterns. △ Less

Submitted 28 June, 2023; v1 submitted 14 October, 2022; originally announced October 2022.

Comments: 26 pages, 9 figures including supplementary material

Journal ref: Bioinspiration and Biomimetics 18 046020 (2023)

Rapid rhythmic entrainment in bio-inspired central pattern generators

Authors: Alex Szorkovszky, Frank Veenstra, Kyrre Glette

Abstract: Entrainment of movement to a periodic stimulus is a characteristic intelligent behaviour in humans and an important goal for adaptive robotics. We demonstrate a quadruped central pattern generator (CPG), consisting of modified Matsuoka neurons, that spontaneously adjusts its period of oscillation to that of a periodic input signal. This is done by simple forcing, with the aid of a filtering networ… ▽ More Entrainment of movement to a periodic stimulus is a characteristic intelligent behaviour in humans and an important goal for adaptive robotics. We demonstrate a quadruped central pattern generator (CPG), consisting of modified Matsuoka neurons, that spontaneously adjusts its period of oscillation to that of a periodic input signal. This is done by simple forcing, with the aid of a filtering network as well as a neural model with tonic input-dependent oscillation period. We first use the NSGA3 algorithm to evolve the CPG parameters, using separate fitness functions for period tunability, limb homogeneity and gait stability. Four CPGs, maximizing different weighted averages of the fitness functions, are then selected from the Pareto front and each is used as a basis for optimizing a filter network. Different numbers of neurons are tested for each filter network. We find that period tunability in particular facilitates robust entrainment, that bounding gaits entrain more easily than walking gaits, and that more neurons in the filter network are beneficial for pre-processing input signals. The system that we present can be used in conjunction with sensory feedback to allow low-level adaptive and robust behaviour in walking robots. △ Less

Submitted 3 June, 2022; originally announced June 2022.

Comments: 7 pages, 6 figures. To appear in Proceedings of the IEEE International Joint Conference on Neural Networks 2022

Friend of a friend models of network growth

Authors: Watson Levens, Alex Szorkovszky, David J. T. Sumpter

Abstract: One of the best-known models in network science is preferential attachment. In this model, the probability of attaching to a node depends on the degree of all nodes in the population, and thus depends on global information. In many biological, physical, and social systems, however, interactions between individuals depend only on local information. Here, we investigate a truly local model of networ… ▽ More One of the best-known models in network science is preferential attachment. In this model, the probability of attaching to a node depends on the degree of all nodes in the population, and thus depends on global information. In many biological, physical, and social systems, however, interactions between individuals depend only on local information. Here, we investigate a truly local model of network formation, based on the idea of a friend of a friend, with the following rule: individuals choose one node at random and link to it with probability p, then they choose a neighbour of that node and link with probability q. Our model produces power laws with empirical exponents ranging from 1.5 upwards and clustering co-efficients ranging from 0 up to 0.5 (consistent with many real networks). For small p and q=1, the model produces super-hub networks, and we prove that for p=0 and q=1, the proportion of non-hubs tends to 1 as the network grows. We show that power-law degree distributions, small world clustering and super-hub networks are all outcomes of this, more general, yet conceptually simple model. △ Less

Submitted 20 September, 2022; v1 submitted 21 December, 2021; originally announced December 2021.

Tomography of an optomechanical oscillator via parametrically amplified position measurement

Authors: P. Warszawski, A. Szorkovszky, W. P. Bowen, A. C. Doherty

Abstract: We propose a protocol for quantum state tomography of nonclassical states in optomechanical systems. Using a parametric drive, the procedure overcomes the challenges of weak optomechanical coupling, poor detection efficiency, and thermal noise to enable high efficiency homodyne measurement. Our analysis is based on the analytic description of the generalized measurement that is performed when opto… ▽ More We propose a protocol for quantum state tomography of nonclassical states in optomechanical systems. Using a parametric drive, the procedure overcomes the challenges of weak optomechanical coupling, poor detection efficiency, and thermal noise to enable high efficiency homodyne measurement. Our analysis is based on the analytic description of the generalized measurement that is performed when optomechanical position measurement competes with thermal noise and a parametric drive. The proposed experimental procedure is numerically simulated in realistic parameter regimes, which allows us to show that tomographic reconstruction of otherwise unverifiable nonclassical states is made possible. △ Less

Submitted 23 June, 2020; v1 submitted 20 November, 2018; originally announced November 2018.

Comments: 37 pages, 5 figures, comments welcome. Published version

An efficient method for sorting and selecting for social behaviour

Authors: Alex Szorkovszky, Alexander Kotrschal, James E. Herbert Read, David J. T. Sumpter, Niclas Kolm, Kristiaan Pelckmans

Abstract: In this article we provide a systematic experimental method for sorting animals according to socially relevant traits, without assaying them or even tagging them individually. Instead, they are repeatedly subjected to behavioural assays in groups, between which the group memberships are rearranged, in order to test the effect of many different combinations of individuals on a group-level property… ▽ More In this article we provide a systematic experimental method for sorting animals according to socially relevant traits, without assaying them or even tagging them individually. Instead, they are repeatedly subjected to behavioural assays in groups, between which the group memberships are rearranged, in order to test the effect of many different combinations of individuals on a group-level property or feature. We analyse this method using a general model for the group feature, and simulate a variety of specific cases to track how individuals are sorted in each case. We find that in the case where the members of a group contribute equally to the group feature, the sorting procedure increases the between-group behavioural variation well above what is expected for groups randomly sampled from a population. For a wide class of group feature models, the individual phenotypes are efficiently sorted across the groups and thus become available for further analysis on how individual properties affect group behaviour. We also show that the experimental data can be used to estimate the individual-level repeatability of the underlying traits. △ Less

Submitted 22 February, 2017; v1 submitted 18 February, 2016; originally announced February 2016.

Comments: 16 pages, 3 figures + supplementary information (3 pages)

Mechanical Entanglement via Detuned Parametric Amplification

Authors: Alex Szorkovszky, Aashish A. Clerk, Andrew C. Doherty, Warwick P. Bowen

Abstract: We propose two schemes to generate entanglement between a pair of mechanical oscillators using parametric amplification. In contrast to existing parametric drive-based protocols, both schemes operate in the steady-state. Using a detuned parametric drive to maintain equilibrium and to couple orthogonal quadratures, our approach can be viewed as a two-mode extension of previous proposals for paramet… ▽ More We propose two schemes to generate entanglement between a pair of mechanical oscillators using parametric amplification. In contrast to existing parametric drive-based protocols, both schemes operate in the steady-state. Using a detuned parametric drive to maintain equilibrium and to couple orthogonal quadratures, our approach can be viewed as a two-mode extension of previous proposals for parametric squeezing. We find that robust steady-state entanglement is possible for matched oscillators with well-controlled coupling. In addition, one of the proposed schemes is robust to differences in the damping rates of the two oscillators. △ Less

Submitted 29 April, 2014; v1 submitted 25 February, 2014; originally announced February 2014.

Comments: 13 pages, 2 figures

Detuned Mechanical Parametric Amplification as a Quantum Non-Demolition Measurement

Authors: Alex Szorkovszky, Aashish A. Clerk, Andrew C. Doherty, Warwick P. Bowen

Abstract: Recently it has been demonstrated that the combination of weak-continuous position detection with detuned parametric driving can lead to significant steady-state mechanical squeezing, far beyond the 3 dB limit normally associated with parametric driving. In this work, we show the close connection between this detuned scheme and quantum non-demolition (QND) measurement of a single mechanical quadra… ▽ More Recently it has been demonstrated that the combination of weak-continuous position detection with detuned parametric driving can lead to significant steady-state mechanical squeezing, far beyond the 3 dB limit normally associated with parametric driving. In this work, we show the close connection between this detuned scheme and quantum non-demolition (QND) measurement of a single mechanical quadrature. In particular, we show that applying an experimentally realistic detuned parametric drive to a cavity optomechanical system allows one to effectively realize a QND measurement despite being in the bad-cavity limit. In the limit of strong squeezing, we show that this scheme offers significant advantages over standard backaction evasion, not only by allowing operation in the weak measurement and low efficiency regimes, but also in terms of the purity of the mechanical state. △ Less

Submitted 29 April, 2014; v1 submitted 23 January, 2014; originally announced January 2014.

Comments: 17 pages, 2 figures

Journal ref: New J. Phys. 16, 043023 (2014)

Strong thermomechanical squeezing via weak measurement

Authors: Alex Szorkovszky, George A. Brawley, Andrew C. Doherty, Warwick P. Bowen

Abstract: We experimentally surpass the 3dB limit to steady state parametric squeezing of a mechanical oscillator. The localization of a AFM cantilever, achieved by optimal estimation, is enhanced by up to 6.2 dB in one position quadrature when a detuned parametric drive is used. This squeezing is, in principle, limited only by the oscillator Q-factor. Used on low temperature, high frequency oscillators, th… ▽ More We experimentally surpass the 3dB limit to steady state parametric squeezing of a mechanical oscillator. The localization of a AFM cantilever, achieved by optimal estimation, is enhanced by up to 6.2 dB in one position quadrature when a detuned parametric drive is used. This squeezing is, in principle, limited only by the oscillator Q-factor. Used on low temperature, high frequency oscillators, this technique provides a pathway to achieve robust quantum squeezing below the zero-point motion. Broadly, our results demonstrate that control systems engineering can overcome well established limits in applications of nonlinear processes. Conversely, by localizing the mechanical position to better than the measurement precision of our apparatus, they demonstrate the usefulness of nonlinearities in control applications. △ Less

Submitted 25 January, 2013; v1 submitted 5 October, 2012; originally announced October 2012.

Comments: 6 pages, 5 figures including supplementary info

Journal ref: Phys. Rev. Lett. 110, 184301 (2013)

Position Estimation of a Parametrically Driven Optomechanical System

Authors: Alex Szorkovszky, Andrew C. Doherty, Glen I. Harris, Warwick P. Bowen

Abstract: We study the position estimation of a mechanical oscillator undergoing both detuned parametric amplification and continuous quantum measurement. This model, which can be utilised to produce squeezed states, is applied to a general optoelectromechanical system. Using a stochastic master equation formalism, we derive general formulae for the reduction in position uncertainty of one quadrature of mot… ▽ More We study the position estimation of a mechanical oscillator undergoing both detuned parametric amplification and continuous quantum measurement. This model, which can be utilised to produce squeezed states, is applied to a general optoelectromechanical system. Using a stochastic master equation formalism, we derive general formulae for the reduction in position uncertainty of one quadrature of motion. The filter for extracting the optimal position estimate from the measurement record is derived. We also find that since this scheme does not work far into the back-action dominated regime, implementing resolved-sideband cooling improves the squeezing only marginally. △ Less

Submitted 3 August, 2012; v1 submitted 31 May, 2012; originally announced May 2012.

Comments: 16 pages, 5 figures

Journal ref: New J. Phys. 14 095026 (2012)

Cavity Optomechanical Magnetometer

Authors: S. Forstner, S. Prams, J. Knittel, E. D. van Ooijen, J. D. Swaim, G. I. Harris, A. Szorkovszky, W. P. Bowen, H. Rubinsztein-Dunlop

Abstract: A cavity optomechanical magnetometer is demonstrated where the magnetic field induced expansion of a magnetostrictive material is transduced onto the physical structure of a highly compliant optical microresonator. The resulting motion is read out optically with ultra-high sensitivity. Detecting the magnetostrictive deformation of Terfenol-D with a toroidal whispering gallery mode (TWGM) resonator… ▽ More A cavity optomechanical magnetometer is demonstrated where the magnetic field induced expansion of a magnetostrictive material is transduced onto the physical structure of a highly compliant optical microresonator. The resulting motion is read out optically with ultra-high sensitivity. Detecting the magnetostrictive deformation of Terfenol-D with a toroidal whispering gallery mode (TWGM) resonator a peak sensitivity of 400 nT/Hz^.5 was achieved with theoretical modelling predicting that sensitivities of up to 500 fT/Hz^.5 may be possible. This chip-based magnetometer combines high-sensitivity and large dynamic range with small size and room temperature operation. △ Less

Submitted 21 February, 2012; v1 submitted 11 October, 2011; originally announced October 2011.

Mechanical squeezing via parametric amplification and weak measurement

Authors: Alex Szorkovszky, Andrew C. Doherty, Glen I. Harris, Warwick P. Bowen

Abstract: Nonlinear forces allow motion of a mechanical oscillator to be squeezed below the zero-point motion. Of existing methods, mechanical parametric amplification is relatively accessible, but previously thought to be limited to 3dB of squeezing in the steady state. We consider the effect of applying continuous weak measurement and feedback to this system. If the parametric drive is optimally detuned f… ▽ More Nonlinear forces allow motion of a mechanical oscillator to be squeezed below the zero-point motion. Of existing methods, mechanical parametric amplification is relatively accessible, but previously thought to be limited to 3dB of squeezing in the steady state. We consider the effect of applying continuous weak measurement and feedback to this system. If the parametric drive is optimally detuned from resonance, correlations between the quadratures of motion allow unlimited steady-state squeezing. Compared to back-action evasion, we demonstrate that the measurement strength, temperature and efficiency requirements for quantum squeezing are significantly relaxed. △ Less

Submitted 29 September, 2011; v1 submitted 7 July, 2011; originally announced July 2011.

Comments: 4 pages, 3 figures

Cavity optoelectromechanical regenerative amplification

Authors: Michael A. Taylor, Alex Szorkovszky, Joachim Knittel, Kwan H. Lee, Terry G. McRae, Warwick P. Bowen

Abstract: Cavity optoelectromechanical regenerative amplification is demonstrated. An optical cavity enhances mechanical transduction, allowing sensitive measurement even for heavy oscillators. A 27.3 MHz mechanical mode of a microtoroid was linewidth narrowed to 6.6\pm1.4 mHz, 30 times smaller than previously achieved with radiation pressure driving in such a system. These results may have applications in… ▽ More Cavity optoelectromechanical regenerative amplification is demonstrated. An optical cavity enhances mechanical transduction, allowing sensitive measurement even for heavy oscillators. A 27.3 MHz mechanical mode of a microtoroid was linewidth narrowed to 6.6\pm1.4 mHz, 30 times smaller than previously achieved with radiation pressure driving in such a system. These results may have applications in areas such as ultrasensitive optomechanical mass spectroscopy. △ Less

Submitted 28 June, 2012; v1 submitted 5 July, 2011; originally announced July 2011.

Journal ref: Optics Express, Vol. 20, Issue 12, pp. 12742-12751 (2012)

Enhanced Zeeman splitting in Ga0.25In0.75As quantum point contacts

Authors: T. P. Martin, A. Szorkovszky, A. P. Micolich, A. R. Hamilton, C. A. Marlow, H. Linke, R. P. Taylor, L. Samuelson

Abstract: The strength of the Zeeman splitting induced by an applied magnetic field is an important factor for the realization of spin-resolved transport in mesoscopic devices. We measure the Zeeman splitting for a quantum point contact etched into a Ga0.25In0.75As quantum well, with the field oriented parallel to the transport direction. We observe an enhancement of the Lande g-factor from |g*|=3.8 +/- 0… ▽ More The strength of the Zeeman splitting induced by an applied magnetic field is an important factor for the realization of spin-resolved transport in mesoscopic devices. We measure the Zeeman splitting for a quantum point contact etched into a Ga0.25In0.75As quantum well, with the field oriented parallel to the transport direction. We observe an enhancement of the Lande g-factor from |g*|=3.8 +/- 0.2 for the third subband to |g*|=5.8 +/- 0.6 for the first subband, six times larger than in GaAs. We report subband spacings in excess of 10 meV, which facilitates quantum transport at higher temperatures. △ Less

Submitted 20 June, 2008; v1 submitted 5 May, 2008; originally announced May 2008.

Comments: [Version 2] Revtex4, 11 pages, 3 figures, accepted for publication in Applied Physics Letters

Journal ref: Applied Physics Letters 93, 012105 (2008)