-
Burning RED: Unlocking Subtask-Driven Reinforcement Learning and Risk-Awareness in Average-Reward Markov Decision Processes
Authors:
Juan Sebastian Rojas,
Chi-Guhn Lee
Abstract:
Average-reward Markov decision processes (MDPs) provide a foundational framework for sequential decision-making under uncertainty. However, average-reward MDPs have remained largely unexplored in reinforcement learning (RL) settings, with the majority of RL-based efforts having been allocated to episodic and discounted MDPs. In this work, we study a unique structural property of average-reward MDP…
▽ More
Average-reward Markov decision processes (MDPs) provide a foundational framework for sequential decision-making under uncertainty. However, average-reward MDPs have remained largely unexplored in reinforcement learning (RL) settings, with the majority of RL-based efforts having been allocated to episodic and discounted MDPs. In this work, we study a unique structural property of average-reward MDPs and utilize it to introduce Reward-Extended Differential (or RED) reinforcement learning: a novel RL framework that can be used to effectively and efficiently solve various subtasks simultaneously in the average-reward setting. We introduce a family of RED learning algorithms for prediction and control, including proven-convergent algorithms for the tabular case. We then showcase the power of these algorithms by demonstrating how they can be used to learn a policy that optimizes, for the first time, the well-known conditional value-at-risk (CVaR) risk measure in a fully-online manner, without the use of an explicit bi-level optimization scheme or an augmented state-space.
△ Less
Submitted 19 October, 2024; v1 submitted 14 October, 2024;
originally announced October 2024.
-
Mono-exponential Current Attenuation with Distance across 16 nm Thick Bacteriorhodopsin Multilayers
Authors:
Domenikos Chryssikos,
Jerry A. Fereiro,
Jonathan Rojas,
Sudipta Bera,
Defne Tüzün,
Evanthia Kounoupioti,
Rui N. Pereira,
Christian Pfeiffer,
Ali Khoshouei,
Hendrik Dietz,
Mordechai Sheves,
David Cahen,
Marc Tornow
Abstract:
The remarkable ability of natural proteins to conduct electricity in the dry state over long distances remains largely inexplicable despite intensive research. In some cases, a (weakly) exponential length-attenuation, as in off-resonant tunneling transport, extends to thicknesses even beyond 10 nm. This report deals with such charge transport characteristics observed in self-assembled multilayers…
▽ More
The remarkable ability of natural proteins to conduct electricity in the dry state over long distances remains largely inexplicable despite intensive research. In some cases, a (weakly) exponential length-attenuation, as in off-resonant tunneling transport, extends to thicknesses even beyond 10 nm. This report deals with such charge transport characteristics observed in self-assembled multilayers of the protein bacteriorhodopsin (bR). About 7.5 nm to 15.5 nm thick bR layers were prepared on conductive titanium nitride (TiN) substrates using aminohexylphosphonic acid and poly-diallyl-dimethylammonium electrostatic linkers. Using conical EGaIn top contacts, an intriguing, mono-exponential conductance attenuation as a function of the bR layer thickness with a small attenuation coefficient $β\approx 0.8 \space {\rm nm}^{-1}$ is measured at zero bias. Variable-temperature measurements using evaporated Ti/Au top contacts yield effective energy barriers of about 100 meV from fitting the data to tunneling, hopping, and carrier cascade transport models. The observed temperature-dependence is assigned to the protein-electrode interfaces. The transport length and temperature dependence of the current densities are consistent with tunneling through the protein-protein and protein-electrode interfaces, respectively. Importantly, our results call for new theoretical approaches to find the microscopic mechanism behind the remarkably efficient, long-range electron transport within bR.
△ Less
Submitted 17 August, 2024;
originally announced August 2024.
-
Current rectification via Photosystem I monolayers induced by their orientation on hydrophilic self-assembled monolayers on titanium nitride
Authors:
Jonathan Rojas,
Zhe Wang,
Feng Liu,
Jerry A. Fereiro,
Domenikos Chryssikos,
Thomas Dittrich,
Dario Leister,
David Cahen,
Marc Tornow
Abstract:
Photosystem I (PSI) is a photosynthetic protein which evolved to efficiently transfer electrons through the thylakoid membrane. This remarkable process attracted the attention of the biomolecular electronics community, which aims to study and understand the underlying electronic transport through these proteins by contacting ensembles of PSI with solid-state metallic contacts. This paper extends p…
▽ More
Photosystem I (PSI) is a photosynthetic protein which evolved to efficiently transfer electrons through the thylakoid membrane. This remarkable process attracted the attention of the biomolecular electronics community, which aims to study and understand the underlying electronic transport through these proteins by contacting ensembles of PSI with solid-state metallic contacts. This paper extends published work of immobilizing monolayers of PSI with a specific orientation, by using organophosphonate self-assembled molecules with hydrophilic heads on ultra-flat titanium nitride. Electrical measurements carried out with eutectic GaIn top contacts showed current rectification ratios of up to ~200. The previously proposed rectification mechanism, relying on the protein's internal electric dipole, was inquired by measuring shifts in the work function. Our straightforward bottom-up fabrication method may allow for further experimental studies on PSI molecules, such as embedding them in solid-state, transparent top contact schemes for optoelectronic measurements.
△ Less
Submitted 17 August, 2024;
originally announced August 2024.
-
On the BCSS Proof of the Fundamental Theorem of Algebra
Authors:
J. Maurice Rojas
Abstract:
Section 10.4 of the 1998 Springer-Verlag book {\em Complexity and Real Computation}, by Blum, Cucker, Shub, and Smale, contains a particularly elegant proof of the Fundamental Theorem of Algebra: The central idea of the proof naturally leads to a homotopy continuation algorithm for finding the roots of univariate polynomials, and extends naturally to a proof of Bézout's Theorem (on the number of r…
▽ More
Section 10.4 of the 1998 Springer-Verlag book {\em Complexity and Real Computation}, by Blum, Cucker, Shub, and Smale, contains a particularly elegant proof of the Fundamental Theorem of Algebra: The central idea of the proof naturally leads to a homotopy continuation algorithm for finding the roots of univariate polynomials, and extends naturally to a proof of Bézout's Theorem (on the number of roots of systems of $n$ equations in $n$ unknowns). We present a more detailed version of the BCSS Proof which is hopefully useful for students and researchers not familiar with algebraic geometry. So while there are no new results in this paper, the exposition is arguably more elementary. Any errors here are solely the responsibility of the current author.
△ Less
Submitted 19 June, 2024; v1 submitted 17 June, 2024;
originally announced June 2024.
-
World Models for General Surgical Grasping
Authors:
Hongbin Lin,
Bin Li,
Chun Wai Wong,
Juan Rojas,
Xiangyu Chu,
Kwok Wai Samuel Au
Abstract:
Intelligent vision control systems for surgical robots should adapt to unknown and diverse objects while being robust to system disturbances. Previous methods did not meet these requirements due to mainly relying on pose estimation and feature tracking. We propose a world-model-based deep reinforcement learning framework "Grasp Anything for Surgery" (GAS), that learns a pixel-level visuomotor poli…
▽ More
Intelligent vision control systems for surgical robots should adapt to unknown and diverse objects while being robust to system disturbances. Previous methods did not meet these requirements due to mainly relying on pose estimation and feature tracking. We propose a world-model-based deep reinforcement learning framework "Grasp Anything for Surgery" (GAS), that learns a pixel-level visuomotor policy for surgical grasping, enhancing both generality and robustness. In particular, a novel method is proposed to estimate the values and uncertainties of depth pixels for a rigid-link object's inaccurate region based on the empirical prior of the object's size; both depth and mask images of task objects are encoded to a single compact 3-channel image (size: 64x64x3) by dynamically zooming in the mask regions, minimizing the information loss. The learned controller's effectiveness is extensively evaluated in simulation and in a real robot. Our learned visuomotor policy handles: i) unseen objects, including 5 types of target grasping objects and a robot gripper, in unstructured real-world surgery environments, and ii) disturbances in perception and control. Note that we are the first work to achieve a unified surgical control system that grasps diverse surgical objects using different robot grippers on real robots in complex surgery scenes (average success rate: 69%). Our system also demonstrates significant robustness across 6 conditions including background variation, target disturbance, camera pose variation, kinematic control error, image noise, and re-grasping after the gripped target object drops from the gripper. Videos and codes can be found on our project page: https://linhongbin.github.io/gas/.
△ Less
Submitted 28 May, 2024;
originally announced May 2024.
-
Bounding Random Test Set Size with Computational Learning Theory
Authors:
Neil Walkinshaw,
Michael Foster,
Jose Miguel Rojas,
Robert M Hierons
Abstract:
Random testing approaches work by generating inputs at random, or by selecting inputs randomly from some pre-defined operational profile. One long-standing question that arises in this and other testing contexts is as follows: When can we stop testing? At what point can we be certain that executing further tests in this manner will not explore previously untested (and potentially buggy) software b…
▽ More
Random testing approaches work by generating inputs at random, or by selecting inputs randomly from some pre-defined operational profile. One long-standing question that arises in this and other testing contexts is as follows: When can we stop testing? At what point can we be certain that executing further tests in this manner will not explore previously untested (and potentially buggy) software behaviors? This is analogous to the question in Machine Learning, of how many training examples are required in order to infer an accurate model. In this paper we show how probabilistic approaches to answer this question in Machine Learning (arising from Computational Learning Theory) can be applied in our testing context. This enables us to produce an upper bound on the number of tests that are required to achieve a given level of adequacy. We are the first to enable this from only knowing the number of coverage targets (e.g. lines of code) in the source code, without needing to observe a sample test executions. We validate this bound on a large set of Java units, and an autonomous driving system.
△ Less
Submitted 24 June, 2024; v1 submitted 27 May, 2024;
originally announced May 2024.
-
Modulated instabilities and the AdS$_2$ point in dense holographic matter
Authors:
Jesús Cruz Rojas,
Tuna Demircik,
Matti Järvinen
Abstract:
We investigate fluctuations of hot and dense QCD plasma by using the gauge/gravity duality. To this end, we carry out a comprehensive classification and analysis of quasinormal modes of charged black holes in the holographic V-QCD model. It turns out that the Chern-Simons term determined by the flavor anomalies of QCD is strong enough to drive a modulated instability. While such an instability is…
▽ More
We investigate fluctuations of hot and dense QCD plasma by using the gauge/gravity duality. To this end, we carry out a comprehensive classification and analysis of quasinormal modes of charged black holes in the holographic V-QCD model. It turns out that the Chern-Simons term determined by the flavor anomalies of QCD is strong enough to drive a modulated instability. While such an instability is expected at high densities, we find that the unstable phase extends to surprisingly low densities and high temperatures, close to the region where data from lattice simulations is available. We also analyze the limit of small temperature which is controlled by a quantum critical AdS$_2$ point. We study in detail the signatures of the critical point in the quasinormal mode spectrum, focusing on the interplay between the hydrodynamic modes and other modes.
△ Less
Submitted 3 May, 2024;
originally announced May 2024.
-
Maximal Number of Skew Lines in the Fermat Surface
Authors:
Sally Andria,
Jacqueline Rojas,
Wállace Mangueira
Abstract:
It is well-known that the Fermat surface of degree $d\geq 3$ has $3d^2$ lines. However, it has not yet been established what is the maximum number of pairwise disjoint lines that it can have if $d\geq 4$. In this article we show that the maximal number of skew lines on the Fermat surface of degree $d \geq 4$ is $3d$ either $d$ even or $d$ odd distinct of $5$, otherwise ($d = 5$) it contains no mor…
▽ More
It is well-known that the Fermat surface of degree $d\geq 3$ has $3d^2$ lines. However, it has not yet been established what is the maximum number of pairwise disjoint lines that it can have if $d\geq 4$. In this article we show that the maximal number of skew lines on the Fermat surface of degree $d \geq 4$ is $3d$ either $d$ even or $d$ odd distinct of $5$, otherwise ($d = 5$) it contains no more than $13$ pairwise disjoint lines.
△ Less
Submitted 22 March, 2024;
originally announced March 2024.
-
Viro's patchworking and the signed reduced A-discriminant
Authors:
Weixun Deng,
J. Maurice Rojas,
Máté L. Telek
Abstract:
Computing the isotopy type of a hypersurface, defined as the positive real zero set of a multivariate polynomial, is a challenging problem in real algebraic geometry. We focus on the case where the defining polynomial has combinatorially restricted exponent vectors and fixed coefficient signs, enabling faster computation of the isotopy type. In particular, Viro's patchworking provides a polyhedral…
▽ More
Computing the isotopy type of a hypersurface, defined as the positive real zero set of a multivariate polynomial, is a challenging problem in real algebraic geometry. We focus on the case where the defining polynomial has combinatorially restricted exponent vectors and fixed coefficient signs, enabling faster computation of the isotopy type. In particular, Viro's patchworking provides a polyhedral complex that has the same isotopy type as the hypersurface, for certain choices of the coefficients. So we present properties of the signed support, focussing mainly on the case of n-variate (n+3)-nomials, that ensure all possible isotopy types can be obtained via patchworking. To prove this, we study the signed reduced A-discriminant and show that it has a simple structure if the signed support satisfies some combinatorial conditions.
△ Less
Submitted 13 March, 2024;
originally announced March 2024.
-
The long-term steady motion of Saturn's Hexagon and the stability of its enclosed jet-stream under seasonal changes
Authors:
A. Sánchez-Lavega,
T. del Río-Gaztelurrutia,
R. Hueso,
S. Pérez-Hoyos,
E. García-Melendo,
A. Antuñano,
I. Mendikoa,
J. F. Rojas,
J. Lillo,
D. Barrado-Navascués,
J. M. Gomez-Forrellad,
C. Go,
D. Peach,
T. Barry,
D. P. Milika,
P. Nicholas,
A. Wesley,
the IOPW-PVOL Team
Abstract:
We investigate the long-term motion of Saturn's North-Pole Hexagon and the structure of its associated eastward jet, using Cassini ISS and ground-based images from 2008 to 2014. We show that both are persistent features that have survived the long polar night, the jet profile remaining essentially unchanged. During those years the hexagon vertices showed a steady rotation period of 10 hr 39 min 23…
▽ More
We investigate the long-term motion of Saturn's North-Pole Hexagon and the structure of its associated eastward jet, using Cassini ISS and ground-based images from 2008 to 2014. We show that both are persistent features that have survived the long polar night, the jet profile remaining essentially unchanged. During those years the hexagon vertices showed a steady rotation period of 10 hr 39 min 23.01 $\pm$ 0.01 s. Analysis of Voyager 1 and 2 (1980-1981) and HST and ground-based (1990-91) images shows a period shorter by 3.5s, due to the presence at the time of a large anticyclone. We interpret the hexagon as a manifestation of a vertically trapped Rossby wave on the polar jet and, because of their survival and unchanged properties under the strong seasonal variations in insolation, we propose that both hexagon and jet are deep-rooted atmospheric features that could reveal the true rotation of the planet Saturn.
△ Less
Submitted 9 February, 2024;
originally announced February 2024.
-
Estimate for the bulk viscosity of strongly coupled quark matter using perturbative QCD and holography
Authors:
Jesús Cruz Rojas,
Tyler Gorda,
Carlos Hoyos,
Niko Jokela,
Matti Järvinen,
Aleksi Kurkela,
Risto Paatelainen,
Saga Säppi,
Aleksi Vuorinen
Abstract:
Modern hydrodynamic simulations of core-collapse supernovae and neutron-star mergers require knowledge not only of the equilibrium properties of strongly interacting matter, but also of the system's response to perturbations, encoded in various transport coefficients. Using perturbative and holographic tools, we derive here an improved weak-coupling and a new strong-coupling result for the most im…
▽ More
Modern hydrodynamic simulations of core-collapse supernovae and neutron-star mergers require knowledge not only of the equilibrium properties of strongly interacting matter, but also of the system's response to perturbations, encoded in various transport coefficients. Using perturbative and holographic tools, we derive here an improved weak-coupling and a new strong-coupling result for the most important transport coefficient of unpaired quark matter, its bulk viscosity. These results are combined in a simple analytic pocket formula for the quantity that is rooted in perturbative Quantum Chromodynamics at high densities but takes into account nonperturbative holographic input at neutron-star densities, where the system is strongly coupled. This expression can be used in the modeling of unpaired quark matter at astrophysically relevant temperatures and densities.
△ Less
Submitted 30 September, 2024; v1 submitted 1 February, 2024;
originally announced February 2024.
-
An Enduring Rapidly Moving Storm as a Guide to Saturn's Equatorial Jet's Complex Structure
Authors:
A. Sánchez-Lavega,
E. García-Melendo,
S. Perez-Hoyos,
R. Hueso,
M. H. Wong,
A. Simon,
J. F. Sanz-Requena,
A. Antuñano,
N. Barrado-Izagirre,
I. Garate-Lopez,
J. F. Rojas,
T. del Rio Gaztelurrutia,
J. M. Gómez-Forrellad,
I. de Pater,
L. Li,
T. Barry,
PVOL contributors
Abstract:
Saturn has an intense and broad eastward equatorial jet with a complex three-dimensional structure mixed with time variability. The equatorial region experiences strong seasonal insolation variations enhanced by ring shadowing and three of the six known giant planetary-scale storms have developed in it. These factors make Saturn's equator a natural laboratory to test models of jets in giant planet…
▽ More
Saturn has an intense and broad eastward equatorial jet with a complex three-dimensional structure mixed with time variability. The equatorial region experiences strong seasonal insolation variations enhanced by ring shadowing and three of the six known giant planetary-scale storms have developed in it. These factors make Saturn's equator a natural laboratory to test models of jets in giant planets. Here we report on a bright equatorial atmospheric feature imaged in 2015 that moved steadily at a high speed of 450 ms-1 not measured since 1980-81 with other equatorial clouds moving within an ample range of velocities. Radiative transfer models show that these motions occur at three altitude levels within the upper haze and clouds. We find that the peak of the jet (latitudes 10\degree N to 10\degree S) suffers intense vertical shears reaching +2.5 ms-1 km-1, two orders of magnitude higher than meridional shears, and temporal variability above 1 bar altitude level.
△ Less
Submitted 31 January, 2024;
originally announced January 2024.
-
Temperature fluctuations in a relativistic gas: Pressure corrections and possible consequences in the deconfinement transition
Authors:
Jorge David Castaño-Yepes,
Marcelo Loewe,
Enrique Muñoz,
Juan Cristóbal Rojas
Abstract:
In this work, we study the effects of random temperature fluctuations on the equation of state of a non-interacting, relativistic fermion gas by means of the replica method. This picture provides a conceptual model for a non-equilibrium system, depicted as an ensemble of subsystems at different temperatures, randomly distributed with respect to a given mean value. We then assume the temperature di…
▽ More
In this work, we study the effects of random temperature fluctuations on the equation of state of a non-interacting, relativistic fermion gas by means of the replica method. This picture provides a conceptual model for a non-equilibrium system, depicted as an ensemble of subsystems at different temperatures, randomly distributed with respect to a given mean value. We then assume the temperature displays stochastic fluctuations $T = T_0 + δT$ with respect to its ensemble average value $T_0$, with zero mean $\overline{δT} = 0$ and standard deviation $\overline{δT^2} = Δ$. By means of the replica method, we obtain the average grand canonical potential, leading to the equation of state of the fermion gas expressed in terms of the excess pressure caused by these fluctuations with respect to the ideal gas at uniform temperature. We further extend our results for the ideal Bose gas as well. Our findings reveal an increase in pressure as the system's ensemble average temperature $T_0$ rises, consistently exceeding the pressure observed in an equilibrium state. Finally, we explore the implications for the deconfinement transition in the context of the simple Bag model, where we show that the critical temperature decreases.
△ Less
Submitted 30 December, 2023;
originally announced January 2024.
-
Spectroradiometry with sub-microsecond time resolution using multianode photomultiplier tube assemblies
Authors:
Zachary M. Geballe,
Francesca Miozzi,
Chris F. Anto,
Javier Rojas,
Jing Yang,
Michael J. Walter
Abstract:
Accurate and precise measurements of spectroradiometric temperature are crucial for many high pressure experiments that use diamond anvil cells or shock waves. In experiments with sub-millisecond timescales, specialized detectors such as streak cameras or photomultiplier tubes are required to measure temperature. High accuracy and precision are difficult to attain, especially at temperatures below…
▽ More
Accurate and precise measurements of spectroradiometric temperature are crucial for many high pressure experiments that use diamond anvil cells or shock waves. In experiments with sub-millisecond timescales, specialized detectors such as streak cameras or photomultiplier tubes are required to measure temperature. High accuracy and precision are difficult to attain, especially at temperatures below 3000 K. Here we present a new spectroradiometry system based on multianode photomultiplier tube technology and passive readout circuitry that yields a 0.24 $μ$s rise-time for each channel. Temperature is measured using five color spectroradiometry. During high pressure pulsed Joule heating experiments in a diamond anvil cell, we document measurement precision to be $\pm 30$ K at temperatures as low as 2000 K during single-shot heating experiments with $0.6$ $μ$s time-resolution. Ambient pressure melting tests using pulsed Joule heating indicate that the accuracy is $\pm 80$ K in the temperature range 1800-2700 K.
△ Less
Submitted 25 October, 2023;
originally announced October 2023.
-
QED Fermions in a noisy magnetic field background: The effective action approach
Authors:
Jorge David Castaño-Yepes,
Marcelo Loewe,
Enrique Muñoz,
Juan Cristóbal Rojas
Abstract:
We consider the effects of a noisy magnetic field background over the fermion propagator in QED, as an approximation to the spatial inhomogeneities and time-fluctuations that would naturally arise in certain physical scenarios, such as heavy-ion collisions or the quark-gluon plasma in the early stages of the evolution of the Universe. We considered a classical, finite and uniform average magnetic…
▽ More
We consider the effects of a noisy magnetic field background over the fermion propagator in QED, as an approximation to the spatial inhomogeneities and time-fluctuations that would naturally arise in certain physical scenarios, such as heavy-ion collisions or the quark-gluon plasma in the early stages of the evolution of the Universe. We considered a classical, finite and uniform average magnetic field background $\langle \mathbf{B}(\mathbf{x})\rangle_Δ = \mathbf{B}$, subject to white-noise fluctuations with auto-correlation of magnitude $Δ_B$. By means of the Schwinger representation of the propagator in the average magnetic field as a reference system, we used the replica formalism to study the effects of the magnetic noise at the mean field level, in terms of a vector order parameter $Q_j = \ii e Δ_B\langle\langle \barψγ_jψ\rangle\rangle_Δ$ whose magnitude represents the ensemble average (over magnetic noise) of the fermion currents. We identified the region where this order parameter acquires a finite value, thus breaking the $U(1)$-symmetry of the model due to the presence of the magnetic noise.
△ Less
Submitted 23 August, 2023;
originally announced August 2023.
-
Drift Rates of Major Neptunian Features between 2018 and 2021
Authors:
Erandi Chavez,
Erin Redwing,
Imke de Pater,
Ricardo Hueso,
Edward M. Molter,
Michael H. Wong,
Carlos Alvarez,
Elinor Gates,
Katherine de Kleer,
Joel Aycock,
Jason Mcilroy,
John Pelletier,
Anthony Ridenour,
Agustín Sánchez-Lavega,
Jose Félix Rojas,
Terry Stickel
Abstract:
Using near-infrared observations of Neptune from the Keck and Lick Observatories, and the Hubble Space Telescope in combination with amateur datasets, we calculated the drift rates of prominent infrared-bright cloud features on Neptune between 2018 and 2021. These features had lifespans of $\sim 1$ day to $\geq$1 month and were located at mid-latitudes and near the south pole. Our observations per…
▽ More
Using near-infrared observations of Neptune from the Keck and Lick Observatories, and the Hubble Space Telescope in combination with amateur datasets, we calculated the drift rates of prominent infrared-bright cloud features on Neptune between 2018 and 2021. These features had lifespans of $\sim 1$ day to $\geq$1 month and were located at mid-latitudes and near the south pole. Our observations permitted determination of drift rates via feature tracking. These drift rates were compared to three zonal wind profiles describing Neptune's atmosphere determined from features tracked in H band (1.6 $μm$), K' band (2.1 $μm$), and Voyager 2 data at visible wavelengths. Features near $-70 °$ measured in the F845M filter (845nm) were particularly consistent with the K' wind profile. The southern mid-latitudes hosted multiple features whose lifespans were $\geq$1 month, providing evidence that these latitudes are a region of high stability in Neptune's atmosphere. We also used HST F467M (467nm) data to analyze a dark, circumpolar wave at $- 60 °$ latitude observed on Neptune since the Voyager 2 era. Its drift rate in recent years (2019-2021) is $4.866 \pm 0.009 °$/day. This is consistent with previous measurements by Karkoschka (2011), which predict a $4.858 \pm 0.022 °$/day drift rate during these years. It also gained a complementary bright band just to the north.
△ Less
Submitted 12 July, 2023;
originally announced July 2023.
-
On the classification of symplectic DQ-algebroids
Authors:
Paul Bressler,
Juan Diego Rojas
Abstract:
DQ-algebroids locally defined on a symplectic manifold form a 2-gerbe. By adapting the method of P. Deligne to the setting of DQ-algebroids we show that this 2-gerbe admits a canonical global section, namely that every symplectic manifold admits a canonical DQ-algebroid quantizing the structure sheaf. The construction relies on methods of non-abelian cohomology and local computations in the Weyl a…
▽ More
DQ-algebroids locally defined on a symplectic manifold form a 2-gerbe. By adapting the method of P. Deligne to the setting of DQ-algebroids we show that this 2-gerbe admits a canonical global section, namely that every symplectic manifold admits a canonical DQ-algebroid quantizing the structure sheaf. The construction relies on methods of non-abelian cohomology and local computations in the Weyl algebra. As a corollary we obtain a classification of symplectic DQ-algebroids.
△ Less
Submitted 6 April, 2023;
originally announced April 2023.
-
Effect of X-Ray Irradiation on Magnetocaloric Material, (MnNiSi)1-x(Fe2Ge)x
Authors:
John Peter J. Nunez,
Vaibhav Sharma,
Jessika V. Rojas,
Radhika Barua,
Ravi L. Hadimani
Abstract:
(MnNiSi)1-(Fe2Ge)x composition (x=0.34) alloy was prepared by arc melting, crushed, and sieved to approximately <32 microns. They were utilized in examining the possible magnetic and structural changes when exposed to a dosage of a continuous sweeping rate of ~>120 Gy/min and an absorbed dose of 35 kGy of X-ray radiation. This study reports observable trends in X-ray diffraction and magnetic measu…
▽ More
(MnNiSi)1-(Fe2Ge)x composition (x=0.34) alloy was prepared by arc melting, crushed, and sieved to approximately <32 microns. They were utilized in examining the possible magnetic and structural changes when exposed to a dosage of a continuous sweeping rate of ~>120 Gy/min and an absorbed dose of 35 kGy of X-ray radiation. This study reports observable trends in X-ray diffraction and magnetic measurements. Magnetization in the magnetization vs. temperature (both heating and cooling) measurements showed an increase from 2.72 emu/g to 4.01 emu/g in the irradiated sample. The Magnetization vs. magnetic field loops exhibited irradiation-induced magnetic hysteresis. The irradiated sample also exhibited an observable change in its coercivity of ΔHc = 14.7% at 200 K. A maximum entropy change ΔSmag of ~ 11.139 J/kg*K and a Tave peak of 317.5 K was achieved for the pristine sample in comparison to ΔSmag of ~ 11.349 J/kg*K and a Tave peak of 312.5 K for the irradiated sample. These presented results provided deeper insights into tuning the effect of irradiation to the magnetic properties of (MnNiSi)1-(Fe2Ge)x for composition (x=0.34) that can be utilized in a wide range of magnetocaloric applications in high-energy radiation environments. The irradiation applied to the sample did not induce any structural or magnetic phase changes in the selected composition but modified the magnetic properties marginally.
△ Less
Submitted 15 March, 2023;
originally announced March 2023.
-
Holographic light dilaton at the conformal edge
Authors:
Jesús Cruz Rojas,
Deog Ki Hong,
Sang Hui Im,
Matti Järvinen
Abstract:
We study a simple holographic model for gauge theories near the conformal edge to show that the dilaton can be parametrically lighter than any other composite states. The masses of all composite states, except the Nambu-Goldstone bosons like dilaton, are bounded by the infrared scale or the dynamical mass. The parametric dependence of the dilaton mass is controlled by the closeness of the anomalou…
▽ More
We study a simple holographic model for gauge theories near the conformal edge to show that the dilaton can be parametrically lighter than any other composite states. The masses of all composite states, except the Nambu-Goldstone bosons like dilaton, are bounded by the infrared scale or the dynamical mass. The parametric dependence of the dilaton mass is controlled by the closeness of the anomalous dimension of the quark bilinear, that breaks spontaneously the scale symmetry, to the conformality. We also show in the holographic dual that under certain assumptions, the dilaton saturates at low energy the anomalous Ward identity for the dilatation currents.
△ Less
Submitted 24 May, 2023; v1 submitted 16 February, 2023;
originally announced February 2023.
-
Popcorn transitions and approach to conformality in homogeneous holographic nuclear matter
Authors:
Jesús Cruz Rojas,
Tuna Demircik,
Matti Järvinen
Abstract:
We study cold and dense nuclear matter by using the gauge/gravity duality. To this end we use the Witten-Sakai-Sugimoto model and the V-QCD models with an approach where the nuclear matter is taken to be spatially homogeneous. We focus on the ''popcorn'' transitions, which are phase transitions in the nuclear matter phases induced by changes in the layer structure of the configuration on the gravi…
▽ More
We study cold and dense nuclear matter by using the gauge/gravity duality. To this end we use the Witten-Sakai-Sugimoto model and the V-QCD models with an approach where the nuclear matter is taken to be spatially homogeneous. We focus on the ''popcorn'' transitions, which are phase transitions in the nuclear matter phases induced by changes in the layer structure of the configuration on the gravity side. We demonstrate that the equation of state for the homogeneous nuclear matter becomes approximately conformal at high densities, and compare our results to other approaches.
△ Less
Submitted 9 January, 2023;
originally announced January 2023.
-
QED Fermions in a noisy magnetic field background
Authors:
Jorge David Castaño-Yepes,
Marcelo Loewe,
Enrique Muñoz,
Juan C. Rojas,
Renato Zamora
Abstract:
We consider the effects of a noisy magnetic field background over the fermion propagator in QED, as an approximation to the spatial inhomogeneities that would naturally arise in certain physical scenarios, such as heavy-ion collisions or the quark-gluon plasma in the early stages of the evolution of the Universe. We considered a classical, finite and uniform average magnetic field background…
▽ More
We consider the effects of a noisy magnetic field background over the fermion propagator in QED, as an approximation to the spatial inhomogeneities that would naturally arise in certain physical scenarios, such as heavy-ion collisions or the quark-gluon plasma in the early stages of the evolution of the Universe. We considered a classical, finite and uniform average magnetic field background $\langle\mathbf{B}(\mathbf{x})\rangle = \mathbf{B}$, subject to white-noise spatial fluctuations with auto-correlation of magnitude $Δ_B$. By means of the Schwinger representation of the propagator in the average magnetic field as a reference system, we used the replica formalism to study the effects of the magnetic noise in the form of renormalized quasi-particle parameters, leading to an effective charge and an effective refraction index, that depend not only on the energy scale, as usual, but also on the magnitude of the noise $Δ_B$ and the average field $\mathbf{B}$.
△ Less
Submitted 30 November, 2022;
originally announced November 2022.
-
A two dimensional fluid model for TCP/AQM analysis
Authors:
Sadek Belamfedel Alaoui,
Alejandro J. Rojas,
Abdelaziz Hmamed,
El Houssaine Tissir
Abstract:
This work proposes a new mathematical model for the TCP/AQM system that aims to improve the accuracy of existing fluid models, especially with respect to the sequential events that occur in the network. The analysis is based on the consideration of two time bases, one at the queue's router level and the other at the congestion window level, which leads to the derivation of a new nonlinear two-dime…
▽ More
This work proposes a new mathematical model for the TCP/AQM system that aims to improve the accuracy of existing fluid models, especially with respect to the sequential events that occur in the network. The analysis is based on the consideration of two time bases, one at the queue's router level and the other at the congestion window level, which leads to the derivation of a new nonlinear two-dimensional fluid model for Internet congestion control. To avoid the difficult task of assessing stability of a 2D nonlinear dynamic model, we perform a local stability analysis of a 2D linear TCP AQM model. By constructing a new two dimensional second order Bessel Legendre Lyapunov functional, new matrix inequalities are derived to evaluate the stability of the 0-input system and to synthesize a feedback controller. Finally, two Internet traffic scenarios, with state space matrices for replicability, are presented, demonstrating the validity of the theoretical results.
△ Less
Submitted 19 November, 2022;
originally announced November 2022.
-
Sturm's Theorem with Endpoints
Authors:
Philippe Pébay,
J. Maurice Rojas,
David C. Thompson
Abstract:
Sturm's Theorem is a fundamental 19th century result relating the number of real roots of a polynomial $f$ in an interval to the number of sign alternations in a sequence of polynomial division-like calculations. We provide a short direct proof of Sturm's Theorem, including the numerically vexing case (ignored in many published accounts) where an interval endpoint is a root of $f$.
Sturm's Theorem is a fundamental 19th century result relating the number of real roots of a polynomial $f$ in an interval to the number of sign alternations in a sequence of polynomial division-like calculations. We provide a short direct proof of Sturm's Theorem, including the numerically vexing case (ignored in many published accounts) where an interval endpoint is a root of $f$.
△ Less
Submitted 16 August, 2022;
originally announced August 2022.
-
Analysis of axial waves in visco-elastic complex structural-acoustic systems: Theory and experiment
Authors:
J. A. Rojas,
A. Morales,
L. Gutiérrez,
J. A. Otero,
E. A. Carrillo,
G. Monsivais,
J. Flores
Abstract:
An experimental and theoretical study of the spectral response of coupled visco-elastic bars subject to axial oscillations is done. Novel closed formulas for the envelope function and their width is derived. These formulas explicitly show the role played by energy dissipation. They show that the internal friction does not affect the width of the envelope of the individual resonances. The formulati…
▽ More
An experimental and theoretical study of the spectral response of coupled visco-elastic bars subject to axial oscillations is done. Novel closed formulas for the envelope function and their width is derived. These formulas explicitly show the role played by energy dissipation. They show that the internal friction does not affect the width of the envelope of the individual resonances. The formulation is based on the equations of classical mechanics combined with Voigt's viscoelastic model. The systems studied consist of a sequence of one, two, or three coupled bars, with their central axes collinear. One of the bars is assumed to be much longer than the others. We discuss the connection between our results with the concept of the strength function phenomenon discovered for the first time in nuclear physics. Our formulation is an alternative and exact approach to the approximated studies based on the fuzzy structure theory that has been used by other authors to describe this type of couplings. The analytical expressions describe the measurements in the laboratory very well.
△ Less
Submitted 12 August, 2022;
originally announced August 2022.
-
Quantum dot molecule devices with optical control of charge status and electronic control of coupling
Authors:
Frederik Bopp,
Jonathan Rojas,
Natalia Revenga,
Hubert Riedl,
Friedrich Sbresny,
Katarina Boos,
Tobias Simmet,
Arash Ahmadi,
David Gershoni,
Jacek Kasprzak,
Arne Ludwig,
Stephan Reitzenstein,
Andreas Wieck,
Dirk Reuter,
Kai Muller,
Jonathan J. Finley
Abstract:
Tunnel-coupled pairs of optically active quantum dots - quantum dot molecules (QDMs) - offer the possibility to combine excellent optical properties such as strong light-matter coupling with two-spin singlet-triplet ($S-T_0$) qubits having extended coherence times. The $S-T_0$ basis formed using two spins is inherently protected against electric and magnetic field noise. However, since a single ga…
▽ More
Tunnel-coupled pairs of optically active quantum dots - quantum dot molecules (QDMs) - offer the possibility to combine excellent optical properties such as strong light-matter coupling with two-spin singlet-triplet ($S-T_0$) qubits having extended coherence times. The $S-T_0$ basis formed using two spins is inherently protected against electric and magnetic field noise. However, since a single gate voltage is typically used to stabilize the charge occupancy of the dots and control the inter-dot orbital couplings, operation of the $S-T_0$ qubits under optimal conditions remains challenging. Here, we present an electric field tunable QDM that can be optically charged with one (1h) or two holes (2h) on demand. We perform a four-phase optical and electric field control sequence that facilitates the sequential preparation of the 2h charge state and subsequently allows flexible control of the inter-dot coupling. Charges are loaded via optical pumping and electron tunnel ionization. We achieve one- and two-hole charging efficiencies of 93.5 $\pm$ 0.8 % and 80.5 $\pm$ 1.3 %, respectively. Combining efficient charge state preparation and precise setting of inter-dot coupling allows control of few-spin qubits, as would be required for on-demand generation of two-dimensional photonic cluster states or quantum transduction between microwaves and photons.
△ Less
Submitted 20 May, 2022;
originally announced May 2022.
-
Trinomials and Deterministic Complexity Limits for Real Solving
Authors:
Erick Boniface,
Weixun Deng,
J. Maurice Rojas
Abstract:
Consider a univariate polynomial f in Z[x] with degree d, exactly t monomial terms, and coefficients in {-H,...,H}. Solving f over the reals, R, in polynomial-time can be defined as counting the exact number of real roots of f and then finding (for each such root z) an approximation w of logarithmic height (log(dH))^{O(1)} such that the Newton iterates of w have error decaying at a rate of O((1/2)…
▽ More
Consider a univariate polynomial f in Z[x] with degree d, exactly t monomial terms, and coefficients in {-H,...,H}. Solving f over the reals, R, in polynomial-time can be defined as counting the exact number of real roots of f and then finding (for each such root z) an approximation w of logarithmic height (log(dH))^{O(1)} such that the Newton iterates of w have error decaying at a rate of O((1/2)^{2^i}). Solving efficiently in this sense, using (log(dH))^{O(1)} deterministic bit operations, is arguably the most honest formulation of solving a polynomial equation over R in time polynomial in the input size. Unfortunately, deterministic algorithms this fast are known only for t=2, unknown for t=3, and provably impossible for t=4. (One can of course resort to older techniques with complexity (d\log H)^{O(1)} for t>=4.)
We give evidence that polynomial-time real-solving in the strong sense above is possible for t=3: We give a polynomial-time algorithm employing A-hypergeometric series that works for all but a fraction of 1/Omega(log(dH)) of the input f. We also show an equivalence between fast trinomial solving and sign evaluation at rational points of small height. As a consequence, we show that for "most" trinomials f, we can compute the sign of f at a rational point r in time polynomial in log(dH) and the logarithmic height of r. (This was known only for binomials before.) We also mention a related family of polynomial systems that should admit a similar speed-up for solving.
△ Less
Submitted 12 February, 2022;
originally announced February 2022.
-
Constraints on the structure and seasonal variations of Triton's atmosphere from the 5 October 2017 stellar occultation and previous observations
Authors:
J. Marques Oliveira,
B. Sicardy,
A. R. Gomes-Júnior,
J. L. Ortiz,
D. F. Strobel,
T. Bertrand,
F. Forget,
E. Lellouch,
J. Desmars,
D. Bérard,
A. Doressoundiram,
J. Lecacheux,
R. Leiva,
E. Meza,
F. Roques,
D. Souami,
T. Widemann,
P. Santos-Sanz,
N. Morales,
R. Duffard,
E. Fernández-Valenzuela,
A. J. Castro-Tirado,
F. Braga-Ribas,
B. E. Morgado,
M. Assafin
, et al. (212 additional authors not shown)
Abstract:
A stellar occultation by Neptune's main satellite, Triton, was observed on 5 October 2017 from Europe, North Africa, and the USA. We derived 90 light curves from this event, 42 of which yielded a central flash detection.
We aimed at constraining Triton's atmospheric structure and the seasonal variations of its atmospheric pressure since the Voyager 2 epoch (1989). We also derived the shape of th…
▽ More
A stellar occultation by Neptune's main satellite, Triton, was observed on 5 October 2017 from Europe, North Africa, and the USA. We derived 90 light curves from this event, 42 of which yielded a central flash detection.
We aimed at constraining Triton's atmospheric structure and the seasonal variations of its atmospheric pressure since the Voyager 2 epoch (1989). We also derived the shape of the lower atmosphere from central flash analysis. We used Abel inversions and direct ray-tracing code to provide the density, pressure, and temperature profiles in the altitude range $\sim$8 km to $\sim$190 km, corresponding to pressure levels from 9 μbar down to a few nanobars.
Results. (i) A pressure of 1.18$\pm$0.03 μbar is found at a reference radius of 1400 km (47 km altitude). (ii) A new analysis of the Voyager 2 radio science occultation shows that this is consistent with an extrapolation of pressure down to the surface pressure obtained in 1989. (iii) A survey of occultations obtained between 1989 and 2017 suggests that an enhancement in surface pressure as reported during the 1990s might be real, but debatable, due to very few high S/N light curves and data accessible for reanalysis. The volatile transport model analysed supports a moderate increase in surface pressure, with a maximum value around 2005-2015 no higher than 23 μbar. The pressures observed in 1995-1997 and 2017 appear mutually inconsistent with the volatile transport model presented here. (iv) The central flash structure does not show evidence of an atmospheric distortion. We find an upper limit of 0.0011 for the apparent oblateness of the atmosphere near the 8 km altitude.
△ Less
Submitted 25 January, 2022;
originally announced January 2022.
-
Efficient color coatings for single junction and multijunction colored solar cells
Authors:
Farid Elsehrawy,
Konrad Klockars,
Orlando J. Rojas,
Janne Halme
Abstract:
Colored solar cells suffer from lower efficiency due to reflection and absorption losses by coatings. By studying different types of coatings on solar cells, the spectrum parameters impacting the solar cell efficiency were identified. A collection of color coatings was fabricated and characterized using spectrophotometry and colorimetric photography. The coatings include commercial absorption filt…
▽ More
Colored solar cells suffer from lower efficiency due to reflection and absorption losses by coatings. By studying different types of coatings on solar cells, the spectrum parameters impacting the solar cell efficiency were identified. A collection of color coatings was fabricated and characterized using spectrophotometry and colorimetric photography. The coatings include commercial absorption filters, commercial distributed bragg reflectors, lab-made interference coatings, and cellulose nanocrystal coatings. Commercial single junction and multijunction solar cells were used to measure the impact of the color coatings on solar cell performance. Structural colors were found to result in the highest brightness to color loss ratio. Structural colors do not fade provided that the structure is not compromised. Color loss in single junction solar cells is a result of the proportion of reflected light in the absorption range of the material. Multijunction solar cells were strongly affected by current mismatch losses, where narrow reflection peaks resulted in high efficiency losses. Structural color coatings made using cellulose nanocrystals exhibited low and broad reflection peaks that produced bright colors and maintained high performance of single junction and multijunction solar cells, retaining up to 83 % of the reference power obtained in the absence of a color coating.
△ Less
Submitted 21 September, 2021;
originally announced September 2021.
-
Root Repulsion and Faster Solving for Very Sparse Polynomials Over $p$-adic Fields
Authors:
J. Maurice Rojas,
Yuyu Zhu
Abstract:
For any fixed field $K\!\in\!\{\mathbb{Q}_2,\mathbb{Q}_3,\mathbb{Q}_5, \ldots\}$, we prove that all polynomials $f\!\in\!\mathbb{Z}[x]$ with exactly $3$ (resp. $2$) monomial terms, degree $d$, and all coefficients having absolute value at most $H$, can be solved over $K$ within deterministic time $\log^{7+o(1)}(dH)$ (resp. $\log^{2+o(1)}(dH)$) in the classical Turing model: Our underlying algorith…
▽ More
For any fixed field $K\!\in\!\{\mathbb{Q}_2,\mathbb{Q}_3,\mathbb{Q}_5, \ldots\}$, we prove that all polynomials $f\!\in\!\mathbb{Z}[x]$ with exactly $3$ (resp. $2$) monomial terms, degree $d$, and all coefficients having absolute value at most $H$, can be solved over $K$ within deterministic time $\log^{7+o(1)}(dH)$ (resp. $\log^{2+o(1)}(dH)$) in the classical Turing model: Our underlying algorithm correctly counts the number of roots of $f$ in $K$, and for each such root generates an approximation in $\mathbb{Q}$ with logarithmic height $O(\log^3(dH))$ that converges at a rate of $O\!\left((1/p)^{2^i}\right)$ after $i$ steps of Newton iteration. We also prove significant speed-ups in certain settings, a minimal spacing bound of $p^{-O(p\log^2_p(dH)\log d)}$ for distinct roots in $\mathbb{C}_p$, and even stronger repulsion when there are nonzero degenerate roots in $\mathbb{C}_p$: $p$-adic distance $p^{-O(\log_p(dH))}$. On the other hand, we prove that there is an explicit family of tetranomials with distinct nonzero roots in $\mathbb{Z}_p$ indistinguishable in their first $Ω(d\log_p H)$ most significant base-$p$ digits.
△ Less
Submitted 19 July, 2021;
originally announced July 2021.
-
Parameter identification for a damage model using a physics informed neural network
Authors:
Carlos J. G. Rojas,
Marco L. Bitterncourt,
José L. Boldrini
Abstract:
This work applies concepts of artificial neural networks to identify the parameters of a mathematical model based on phase fields for damage and fracture. Damage mechanics is the part of the continuum mechanics that models the effects of the micro-defect formation using state variables at the macroscopic level. The equations that define the model are derived from fundamental laws of physics and pr…
▽ More
This work applies concepts of artificial neural networks to identify the parameters of a mathematical model based on phase fields for damage and fracture. Damage mechanics is the part of the continuum mechanics that models the effects of the micro-defect formation using state variables at the macroscopic level. The equations that define the model are derived from fundamental laws of physics and provide important relationships between state variables. Simulations using the model considered in this work produce good qualitative and quantitative results, but many parameters must be adjusted to reproduce a certain material behavior. The identification of model parameters is considered by solving an inverse problem that uses pseudo-experimental data to find the values that produce the best fit to the data. We apply a physics-informed neural network and combine some classical estimation methods to identify the material parameters that appear in the damage equation of the model. Our strategy consists of a neural network that acts as an approximating function of the damage evolution with its output regularized using the residue of the differential equation. Three stages of optimization seek the best possible values for the neural network and the material parameters. The training alternates between the fitting of only the pseudo-experimental data or the total loss that includes the regularizing terms. We test the robustness of the method to noisy data and its generalization capabilities using a simple physical case for the damage model. This procedure deals better with noisy data in comparison with a PDE-constrained optimization method, and it also provides good approximations of the material parameters and the evolution of damage.
△ Less
Submitted 25 June, 2021;
originally announced July 2021.
-
Holographic Colour Superconductors at Finite Coupling with NJL Interactions
Authors:
Kazem Bitaghsir Fadafan,
Jesús Cruz Rojas
Abstract:
We study a bottom-up holographic description of the QCD colour superconducting phase in the presence of higher derivative corrections. We expand this holographic model in the context of Gauss-Bonnet (GB) gravity. The Cooper pair condensate has been investigated in the deconfinement phase for different values of the GB coupling parameter $λ_{G B}$, we observe a change in the value of the critical c…
▽ More
We study a bottom-up holographic description of the QCD colour superconducting phase in the presence of higher derivative corrections. We expand this holographic model in the context of Gauss-Bonnet (GB) gravity. The Cooper pair condensate has been investigated in the deconfinement phase for different values of the GB coupling parameter $λ_{G B}$, we observe a change in the value of the critical chemical potential $μ_c$ in comparison to Einstein gravity. We find that $μ_c$ grows as $λ_{G B}$ increases. We add four fermion interactions and show that in the presence of these corrections the main interesting features of the model are still present and that the intrinsic attractive interaction can not be switched off. This study suggests to find GB corrections to equation of state of holographic QCD matter.
△ Less
Submitted 9 July, 2021;
originally announced July 2021.
-
Domain Wall Fermions on the Brane
Authors:
Jesus Cruz Rojas,
Nick Evans,
Jack Mitchell
Abstract:
We study domain wall fermions and their condensation in the D3/probe D7 system. A spatially dependent mass term for the N=2 hypermultiplet can be arranged to isolate distinct two component fermions on two 2+1 dimensional domain walls. We argue that the system shows condensation/mass generation analogous to the D3/probe D5 anti-D5 system. The chiral condensate and pion mass can be directly computed…
▽ More
We study domain wall fermions and their condensation in the D3/probe D7 system. A spatially dependent mass term for the N=2 hypermultiplet can be arranged to isolate distinct two component fermions on two 2+1 dimensional domain walls. We argue that the system shows condensation/mass generation analogous to the D3/probe D5 anti-D5 system. The chiral condensate and pion mass can be directly computed on the domain wall. We provide evidence that these systems with the domains separated by a width w have a bare (current) quark mass that scales as 1/w when the spatial dependent mass is large. Adding a magnetic field does not induce chiral symmetry breaking between the separated domain wall fermions, but a similar phenomenological dilaton factor can be made strong enough to introduce spontaneous symmetry breaking. We show a Gell-Man-Oakes-Renner relation for the pions in that case and also for the case where the D7 probe is in a back-reacted dilaton flow geometry. The vacuum configurations can also be interpreted as having a spontaneously generated mass by a Nambu-Jona-Lasinio four fermion operator, depending on the choice of boundary conditions on fluctuations, according to Witten's multi-trace prescription.
△ Less
Submitted 16 June, 2021;
originally announced June 2021.
-
Towards Safe Control of Continuum Manipulator Using Shielded Multiagent Reinforcement Learning
Authors:
Guanglin Ji,
Junyan Yan,
Jingxin Du,
Wanquan Yan,
Jibiao Chen,
Yongkang Lu,
Juan Rojas,
Shing Shin Cheng
Abstract:
Continuum robotic manipulators are increasingly adopted in minimal invasive surgery. However, their nonlinear behavior is challenging to model accurately, especially when subject to external interaction, potentially leading to poor control performance. In this letter, we investigate the feasibility of adopting a model-free multiagent reinforcement learning (RL), namely multiagent deep Q network (M…
▽ More
Continuum robotic manipulators are increasingly adopted in minimal invasive surgery. However, their nonlinear behavior is challenging to model accurately, especially when subject to external interaction, potentially leading to poor control performance. In this letter, we investigate the feasibility of adopting a model-free multiagent reinforcement learning (RL), namely multiagent deep Q network (MADQN), to control a 2-degree of freedom (DoF) cable-driven continuum surgical manipulator. The control of the robot is formulated as a one-DoF, one agent problem in the MADQN framework to improve the learning efficiency. Combined with a shielding scheme that enables dynamic variation of the action set boundary, MADQN leads to efficient and importantly safer control of the robot. Shielded MADQN enabled the robot to perform point and trajectory tracking with submillimeter root mean square errors under external loads, soft obstacles, and rigid collision, which are common interaction scenarios encountered by surgical manipulators. The controller was further proven to be effective in a miniature continuum robot with high structural nonlinearitiy, achieving trajectory tracking with submillimeter accuracy under external payload.
△ Less
Submitted 15 June, 2021;
originally announced June 2021.
-
Error Identification and Recovery in Robotic Snap Assembly
Authors:
Yusuke Hayami,
Weiwei Wan,
Keisuke Koyama,
Peihao Shi,
Juan Rojas,
Kensuke Harada
Abstract:
Existing methods for predicting robotic snap joint assembly cannot predict failures before their occurrence. To address this limitation, this paper proposes a method for predicting error states before the occurence of error, thereby enabling timely recovery. Robotic snap joint assembly requires precise positioning; therefore, even a slight offset between parts can lead to assembly failure. To corr…
▽ More
Existing methods for predicting robotic snap joint assembly cannot predict failures before their occurrence. To address this limitation, this paper proposes a method for predicting error states before the occurence of error, thereby enabling timely recovery. Robotic snap joint assembly requires precise positioning; therefore, even a slight offset between parts can lead to assembly failure. To correctly predict error states, we apply functional principal component analysis (fPCA) to 6D force/torque profiles that are terminated before the occurence of an error. The error state is identified by applying a feature vector to a decision tree, wherein the support vector machine (SVM) is employed at each node. If the estimation accuracy is low, we perform additional probing to more correctly identify the error state. Finally, after identifying the error state, a robot performs the error recovery motion based on the identified error state. Through the experimental results of assembling plastic parts with four snap joints, we show that the error states can be correctly estimated and a robot can recover from the identified error state.
△ Less
Submitted 24 March, 2021;
originally announced March 2021.
-
Jupiter's third largest and longest-lived oval: Color changes and dynamics
Authors:
N. Barrado-Izagirre,
J. Legarreta,
A. Sánchez-Lavega,
S. Pérez-Hoyos,
R. Hueso,
P. Iñurrigarro,
J. F. Rojas,
I. Mendikoa,
I. Ordoñez-Etxeberria,
the IOPW Team
Abstract:
The transition region between the North Equatorial Band (NEBn) and North Tropical Zone (NTrZ) in Jupiter is home to convective storms, systems of cyclones and anticyclones and atmospheric waves. A large anticyclone formed in the year 2006 at planetographic latitude 19N and persists since then after a complex dynamic history, being possibly the third longest-lived oval in the planet after Jupiter's…
▽ More
The transition region between the North Equatorial Band (NEBn) and North Tropical Zone (NTrZ) in Jupiter is home to convective storms, systems of cyclones and anticyclones and atmospheric waves. A large anticyclone formed in the year 2006 at planetographic latitude 19N and persists since then after a complex dynamic history, being possibly the third longest-lived oval in the planet after Jupiter's Great Red Spot and oval BA. This anticyclone has experienced close interactions with other ovals, merging with another oval in February 2013; it has also experienced color changes, from white to red (September 2013). The oval survived the effects of the closely located North Temperate Belt Disturbance, which occurred in October 2016 and fully covered the oval, rendering it unobservable for a short time. When it became visible again at its expected longitude from its previous longitudinal track, it reappeared as a white large oval keeping this color and the same morphology since 2017 at least until the onset of the new convective disturbance in Jupiter's North Temperate Belt in August 2020. Here we describe the historic evolution of the properties of this oval. We use JunoCam and Hubble Space Telescope (HST) images to measure its size and its internal rotation. We also used HST and PlanetCam-UPV/EHU multi-wavelength observations to characterize its color changes and Junocam images to unveil its detailed structure. The color and the altitude-opacity indices show that the oval is higher and has redder clouds than its environment but has lower cloud tops than other large ovals like the GRS, and it is less red than the GRS and oval BA. We show that in spite of the dramatic environmental changes suffered by the oval during all these years, its main characteristics are stable in time and therefore must be related with the atmospheric dynamics below the observable cloud decks.
△ Less
Submitted 24 March, 2021;
originally announced March 2021.
-
Water Contribution to the Protein Folding and its Relevance in Protein Design and Protein Aggregation
Authors:
Giancarlo Franzese,
Joan Àguila Rojas,
Valentino Bianco,
Ivan Coluzza
Abstract:
Water plays a fundamental role in protein stability. However, the effect of the properties of water on the behaviour of proteins is only partially understood. Several theories have been proposed to give insight into the mechanisms of cold and pressure denaturation, or the limits of temperature and pressure above which no protein has a stable, functional state, or how unfolding and aggregation are…
▽ More
Water plays a fundamental role in protein stability. However, the effect of the properties of water on the behaviour of proteins is only partially understood. Several theories have been proposed to give insight into the mechanisms of cold and pressure denaturation, or the limits of temperature and pressure above which no protein has a stable, functional state, or how unfolding and aggregation are related. Here we review our results based on a theoretical approach that can rationalise the water contribution to protein solutions' free energy. We show, using Monte Carlo simulations, how we can rationalise experimental data with our recent results. We discuss how our findings can help develop new strategies for the design of novel synthetic biopolymers or possible approaches for mitigating neurodegenerative pathologies.
△ Less
Submitted 24 March, 2021;
originally announced March 2021.
-
Typical coexistence of infinitely many strange attractors
Authors:
Pablo G. Barrientos,
Juan David Rojas
Abstract:
We prove that the coexistence of infinitely many prevalent Hénon-like phenomena is Kolmogorov typical in sectional dissipative $C^{d,r}$-Berger domains of parameter families of diffeomorphisms of dimension $m\geq 3$ for $d<r-1$. Namely, we answer an old question posed by Colli in [Annales de l'Institut Henri Poincare-Nonlinear Analysis, 15, 539--580 (1998)] on typicality of the coexistence of infi…
▽ More
We prove that the coexistence of infinitely many prevalent Hénon-like phenomena is Kolmogorov typical in sectional dissipative $C^{d,r}$-Berger domains of parameter families of diffeomorphisms of dimension $m\geq 3$ for $d<r-1$. Namely, we answer an old question posed by Colli in [Annales de l'Institut Henri Poincare-Nonlinear Analysis, 15, 539--580 (1998)] on typicality of the coexistence of infinitely many non-hyperbolic strange attractors for $3\leq d<r-1$.
△ Less
Submitted 11 January, 2023; v1 submitted 16 February, 2021;
originally announced February 2021.
-
Differentiable Implicit Soft-Body Physics
Authors:
Junior Rojas,
Eftychios Sifakis,
Ladislav Kavan
Abstract:
We present a differentiable soft-body physics simulator that can be composed with neural networks as a differentiable layer. In contrast to other differentiable physics approaches that use explicit forward models to define state transitions, we focus on implicit state transitions defined via function minimization. Implicit state transitions appear in implicit numerical integration methods, which o…
▽ More
We present a differentiable soft-body physics simulator that can be composed with neural networks as a differentiable layer. In contrast to other differentiable physics approaches that use explicit forward models to define state transitions, we focus on implicit state transitions defined via function minimization. Implicit state transitions appear in implicit numerical integration methods, which offer the benefits of large time steps and excellent numerical stability, but require a special treatment to achieve differentiability due to the absence of an explicit differentiable forward pass. In contrast to other implicit differentiation approaches that require explicit formulas for the force function and the force Jacobian matrix, we present an energy-based approach that allows us to compute these derivatives automatically and in a matrix-free fashion via reverse-mode automatic differentiation. This allows for more flexibility and productivity when defining physical models and is particularly important in the context of neural network training, which often relies on reverse-mode automatic differentiation (backpropagation). We demonstrate the effectiveness of our differentiable simulator in policy optimization for locomotion tasks and show that it achieves better sample efficiency than model-free reinforcement learning.
△ Less
Submitted 9 September, 2021; v1 submitted 10 February, 2021;
originally announced February 2021.
-
Reduced-order Model for Fluid Flows via Neural Ordinary Differential Equations
Authors:
Carlos J. G. Rojas,
Andreas Dengel,
Mateus Dias Ribeiro
Abstract:
Reduced order models play an important role in the design, optimization and control of dynamical systems. In recent years, there has been an increasing interest in the application of data-driven techniques for model reduction that can decrease the computational burden of numerical solutions, while preserving the most important features of complex physical problems. In this paper, we use the proper…
▽ More
Reduced order models play an important role in the design, optimization and control of dynamical systems. In recent years, there has been an increasing interest in the application of data-driven techniques for model reduction that can decrease the computational burden of numerical solutions, while preserving the most important features of complex physical problems. In this paper, we use the proper orthogonal decomposition to reduce the dimensionality of the model and introduce a novel generative neural ODE (NODE) architecture to forecast the behavior of the temporal coefficients. With this methodology, we replace the classical Galerkin projection with an architecture characterized by the use of a continuous latent space. We exemplify the methodology on the dynamics of the Von Karman vortex street of the flow past a cylinder generated by a Large-eddy Simulation (LES)-based code. We compare the NODE methodology with an LSTM baseline to assess the extrapolation capabilities of the generative model and present some qualitative evaluations of the flow reconstructions.
△ Less
Submitted 20 February, 2021; v1 submitted 3 February, 2021;
originally announced February 2021.
-
Sub-Linear Point Counting for Variable Separated Curves over Prime Power Rings
Authors:
Caleb Robelle,
J. Maurice Rojas,
Yuyu Zhu
Abstract:
Let $k,p\in \mathbb{N}$ with $p$ prime and let $f\in\mathbb{Z}[x_1,x_2]$ be a bivariate polynomial with degree $d$ and all coefficients of absolute value at most $p^k$. Suppose also that $f$ is variable separated, i.e., $f=g_1+g_2$ for $g_i\in\mathbb{Z}[x_i]$. We give the first algorithm, with complexity sub-linear in $p$, to count the number of roots of $f$ over $\mathbb{Z}$ mod $p^k$ for arbitra…
▽ More
Let $k,p\in \mathbb{N}$ with $p$ prime and let $f\in\mathbb{Z}[x_1,x_2]$ be a bivariate polynomial with degree $d$ and all coefficients of absolute value at most $p^k$. Suppose also that $f$ is variable separated, i.e., $f=g_1+g_2$ for $g_i\in\mathbb{Z}[x_i]$. We give the first algorithm, with complexity sub-linear in $p$, to count the number of roots of $f$ over $\mathbb{Z}$ mod $p^k$ for arbitrary $k$: Our Las Vegas randomized algorithm works in time $(dk\log p)^{O(1)}\sqrt{p}$, and admits a quantum version for smooth curves working in time $(d\log p)^{O(1)}k$. Save for some subtleties concerning non-isolated singularities, our techniques generalize to counting roots of polynomials in $\mathbb{Z}[x_1,\ldots,x_n]$ over $\mathbb{Z}$ mod $p^k$.
Our techniques are a first step toward efficient point counting for varieties over Galois rings (which is relevant to error correcting codes over higher-dimensional varieties), and also imply new speed-ups for computing Igusa zeta functions of curves. The latter zeta functions are fundamental in arithmetic geometry.
△ Less
Submitted 2 February, 2021;
originally announced February 2021.
-
Reconstructing phase-resolved hysteresis loops from first-order reversal curves
Authors:
Dustin A. Gilbert,
Peyton D. Murray,
Julius De Rojas,
Randy K. Dumas,
Joseph E. Davies,
Kai Liu
Abstract:
The first order reversal curve (FORC) method is a magnetometry based technique used to capture nanoscale magnetic phase separation and interactions with macroscopic measurements using minor hysteresis loop analysis. This makes the FORC technique a powerful tool in the analysis of complex systems which cannot be effectively probed using localized techniques. However, recovering quantitative details…
▽ More
The first order reversal curve (FORC) method is a magnetometry based technique used to capture nanoscale magnetic phase separation and interactions with macroscopic measurements using minor hysteresis loop analysis. This makes the FORC technique a powerful tool in the analysis of complex systems which cannot be effectively probed using localized techniques. However, recovering quantitative details about the identified phases which can be compared to traditionally measured metrics remains an enigmatic challenge. We demonstrate a technique to reconstruct phase-resolved magnetic hysteresis loops by selectively integrating the measured FORC distribution. From these minor loops, the traditional metrics - including the coercivity and saturation field, and the remanent and saturation magnetization - can be determined. In order to perform this analysis, special consideration must be paid to the accurate quantitative management of the so-called reversible features. This technique is demonstrated on three representative materials systems, high anisotropy FeCuPt thin-films, Fe nanodots, and SmCo/Fe exchange spring magnet films, and shows excellent agreement with the direct measured major loop, as well as the phase separated loops.
△ Less
Submitted 20 December, 2020;
originally announced December 2020.
-
Counting Real Roots in Polynomial-Time for Systems Supported on Circuits
Authors:
J. Maurice Rojas
Abstract:
Suppose $A=\{a_1,\ldots,a_{n+2}\}\subset\mathbb{Z}^n$ has cardinality $n+2$, with all the coordinates of the $a_j$ having absolute value at most $d$, and the $a_j$ do not all lie in the same affine hyperplane. Suppose $F=(f_1,\ldots,f_n)$ is an $n\times n$ polynomial system with generic integer coefficients at most $H$ in absolute value, and $A$ the union of the sets of exponent vectors of the…
▽ More
Suppose $A=\{a_1,\ldots,a_{n+2}\}\subset\mathbb{Z}^n$ has cardinality $n+2$, with all the coordinates of the $a_j$ having absolute value at most $d$, and the $a_j$ do not all lie in the same affine hyperplane. Suppose $F=(f_1,\ldots,f_n)$ is an $n\times n$ polynomial system with generic integer coefficients at most $H$ in absolute value, and $A$ the union of the sets of exponent vectors of the $f_i$. We give the first algorithm that, for any fixed $n$, counts exactly the number of real roots of $F$ in in time polynomial in $\log(dH)$.
△ Less
Submitted 11 June, 2021; v1 submitted 9 December, 2020;
originally announced December 2020.
-
Hyperparameter Auto-tuning in Self-Supervised Robotic Learning
Authors:
Jiancong Huang,
Juan Rojas,
Matthieu Zimmer,
Hongmin Wu,
Yisheng Guan,
Paul Weng
Abstract:
Policy optimization in reinforcement learning requires the selection of numerous hyperparameters across different environments. Fixing them incorrectly may negatively impact optimization performance leading notably to insufficient or redundant learning. Insufficient learning (due to convergence to local optima) results in under-performing policies whilst redundant learning wastes time and resource…
▽ More
Policy optimization in reinforcement learning requires the selection of numerous hyperparameters across different environments. Fixing them incorrectly may negatively impact optimization performance leading notably to insufficient or redundant learning. Insufficient learning (due to convergence to local optima) results in under-performing policies whilst redundant learning wastes time and resources. The effects are further exacerbated when using single policies to solve multi-task learning problems. Observing that the Evidence Lower Bound (ELBO) used in Variational Auto-Encoders correlates with the diversity of image samples, we propose an auto-tuning technique based on the ELBO for self-supervised reinforcement learning. Our approach can auto-tune three hyperparameters: the replay buffer size, the number of policy gradient updates during each epoch, and the number of exploration steps during each epoch. We use a state-of-the-art self-supervised robot learning framework (Reinforcement Learning with Imagined Goals (RIG) using Soft Actor-Critic) as baseline for experimental verification. Experiments show that our method can auto-tune online and yields the best performance at a fraction of the time and computational resources. Code, video, and appendix for simulated and real-robot experiments can be found at the project page \url{www.JuanRojas.net/autotune}.
△ Less
Submitted 24 March, 2021; v1 submitted 16 October, 2020;
originally announced October 2020.
-
Holographic quark matter with colour superconductivity and a stiff equation of state for compact stars
Authors:
Kazem Bitaghsir Fadafan,
Jesus Cruz Rojas,
Nick Evans
Abstract:
We present a holographic model of QCD with a first order chiral restoration phase transition with chemical potential, mu. The first order behaviour follows from allowing a discontinuity in the dual description as the quarks are integrated out below their constituent mass. The model predicts a deconfined yet massive quark phase at intermediate densities (350 MeV< mu <500 MeV), above the nuclear den…
▽ More
We present a holographic model of QCD with a first order chiral restoration phase transition with chemical potential, mu. The first order behaviour follows from allowing a discontinuity in the dual description as the quarks are integrated out below their constituent mass. The model predicts a deconfined yet massive quark phase at intermediate densities (350 MeV< mu <500 MeV), above the nuclear density phase, which has a very stiff equation of state and a speed of sound close to one. We also include a holographic description of a colour superconducting condensate in the chirally restored vacuum and study the resulting equation of state. They provides a well behaved first order transition from the deconfined massive quark phase at very high density (mu>500 MeV). We solve the Tolman-Oppenheimer-Volkoff equations with the resulting equations of state and find stable hybrid stars with quark cores. We compute the tidal deformability for these hybrid stars and show they are consistent with LIGO/Virgo data on a neutron star collision. Our holographic model shows that quark matter could be present at the core of such compact stars.
△ Less
Submitted 29 September, 2020;
originally announced September 2020.
-
Conformal Defects from String Field Theory
Authors:
Kasia Budzik,
Miroslav Rapcak,
Jairo M. Rojas
Abstract:
Unlike conformal boundary conditions, conformal defects of Virasoro minimal models lack classification. Alternatively to the defect perturbation theory and the truncated conformal space approach, we employ open string field theory (OSFT) techniques to explore the space of conformal defects. We illustrate the method by an analysis of OSFT around the background associated to the $(1,2)$ topological…
▽ More
Unlike conformal boundary conditions, conformal defects of Virasoro minimal models lack classification. Alternatively to the defect perturbation theory and the truncated conformal space approach, we employ open string field theory (OSFT) techniques to explore the space of conformal defects. We illustrate the method by an analysis of OSFT around the background associated to the $(1,2)$ topological defect in diagonal unitary minimal models. Numerical analysis of OSFT equations of motion leads to an identification of a nice family of solutions, recovering the picture of infrared fixed points due to Kormos, Runkel and Watts. In particular, we find a continuum of solutions in the Ising model case and 6 solutions for other minimal models. OSFT provides us with numerical estimates of the g-function and other coefficients of the boundary state.
△ Less
Submitted 23 January, 2021; v1 submitted 14 September, 2020;
originally announced September 2020.
-
Pose-based Modular Network for Human-Object Interaction Detection
Authors:
Zhijun Liang,
Junfa Liu,
Yisheng Guan,
Juan Rojas
Abstract:
Human-object interaction(HOI) detection is a critical task in scene understanding. The goal is to infer the triplet <subject, predicate, object> in a scene. In this work, we note that the human pose itself as well as the relative spatial information of the human pose with respect to the target object can provide informative cues for HOI detection. We contribute a Pose-based Modular Network (PMN) w…
▽ More
Human-object interaction(HOI) detection is a critical task in scene understanding. The goal is to infer the triplet <subject, predicate, object> in a scene. In this work, we note that the human pose itself as well as the relative spatial information of the human pose with respect to the target object can provide informative cues for HOI detection. We contribute a Pose-based Modular Network (PMN) which explores the absolute pose features and relative spatial pose features to improve HOI detection and is fully compatible with existing networks. Our module consists of a branch that first processes the relative spatial pose features of each joint independently. Another branch updates the absolute pose features via fully connected graph structures. The processed pose features are then fed into an action classifier. To evaluate our proposed method, we combine the module with the state-of-the-art model named VS-GATs and obtain significant improvement on two public benchmarks: V-COCO and HICO-DET, which shows its efficacy and flexibility. Code is available at \url{https://github.com/birlrobotics/PMN}.
△ Less
Submitted 5 August, 2020;
originally announced August 2020.
-
Computing zeta functions of large polynomial systems over finite fields
Authors:
Qi Cheng,
J. Maurice Rojas,
Daqing Wan
Abstract:
In this paper, we improve the algorithms of Lauder-Wan \cite{LW} and Harvey \cite{Ha} to compute the zeta function of a system of $m$ polynomial equations in $n$ variables over the finite field $\FF_q$ of $q$ elements, for $m$ large. The dependence on $m$ in the original algorithms was exponential in $m$. Our main result is a reduction of the exponential dependence on $m$ to a polynomial dependenc…
▽ More
In this paper, we improve the algorithms of Lauder-Wan \cite{LW} and Harvey \cite{Ha} to compute the zeta function of a system of $m$ polynomial equations in $n$ variables over the finite field $\FF_q$ of $q$ elements, for $m$ large. The dependence on $m$ in the original algorithms was exponential in $m$. Our main result is a reduction of the exponential dependence on $m$ to a polynomial dependence on $m$. As an application, we speed up a doubly exponential time algorithm from a software verification paper \cite{BJK} (on universal equivalence of programs over finite fields) to singly exponential time. One key new ingredient is an effective version of the classical Kronecker theorem which (set-theoretically) reduces the number of defining equations for a "large" polynomial system over $\FF_q$ when $q$ is suitably large.
△ Less
Submitted 26 July, 2020;
originally announced July 2020.
-
Chiral symmetry restoration in RQED at finite temperature in the supercritical coupling regime
Authors:
Jean Báez,
Alfredo Raya,
J. C. Rojas
Abstract:
We explore the conditions for chiral symmetry breaking in reduced (or pseudo) quantum electrodynamics at finite temperature in connection with graphene and other 2D-materials with an underlying Dirac behavior of the charge carriers. By solving the corresponding Schwinger-Dyson equation in rainbow approximation, in a Landau-like gauge and neglecting wavefunction renormalization effects, we find the…
▽ More
We explore the conditions for chiral symmetry breaking in reduced (or pseudo) quantum electrodynamics at finite temperature in connection with graphene and other 2D-materials with an underlying Dirac behavior of the charge carriers. By solving the corresponding Schwinger-Dyson equation in rainbow approximation, in a Landau-like gauge and neglecting wavefunction renormalization effects, we find the need of the coupling to exceed a critical value $α_c$ in order for chiral symmetry to be broken, in agreement with known results from other groups. In this supercritical regime, we add the effects of a thermal bath at temperature $T$ and find the critical values of this parameter that leads to chiral symmetry restoration.
△ Less
Submitted 7 September, 2020; v1 submitted 10 July, 2020;
originally announced July 2020.
-
Global Field Totients
Authors:
Santiago Arango-Piñeros,
Juan Diego Rojas
Abstract:
Using a remainder theorem for valuations of a field, we give a new perspective on the norm function of a global field. We define the Euler totient function of a global field and recover the essential analytical properties of the classical arithmetical function, namely the product formula. In addition, we prove the holomorphicity of the associated zeta function. As an application, we recover the an…
▽ More
Using a remainder theorem for valuations of a field, we give a new perspective on the norm function of a global field. We define the Euler totient function of a global field and recover the essential analytical properties of the classical arithmetical function, namely the product formula. In addition, we prove the holomorphicity of the associated zeta function. As an application, we recover the analog of the mean value theorem of Erdős, Dressler, and Bateman via the Weiner-Ikehara theorem.
△ Less
Submitted 11 May, 2020; v1 submitted 9 May, 2020;
originally announced May 2020.
-
A Graph Attention Spatio-temporal Convolutional Network for 3D Human Pose Estimation in Video
Authors:
Junfa Liu,
Juan Rojas,
Zhijun Liang,
Yihui Li,
Yisheng Guan
Abstract:
Spatio-temporal information is key to resolve occlusion and depth ambiguity in 3D pose estimation. Previous methods have focused on either temporal contexts or local-to-global architectures that embed fixed-length spatio-temporal information. To date, there have not been effective proposals to simultaneously and flexibly capture varying spatio-temporal sequences and effectively achieves real-time…
▽ More
Spatio-temporal information is key to resolve occlusion and depth ambiguity in 3D pose estimation. Previous methods have focused on either temporal contexts or local-to-global architectures that embed fixed-length spatio-temporal information. To date, there have not been effective proposals to simultaneously and flexibly capture varying spatio-temporal sequences and effectively achieves real-time 3D pose estimation. In this work, we improve the learning of kinematic constraints in the human skeleton: posture, local kinematic connections, and symmetry by modeling local and global spatial information via attention mechanisms. To adapt to single- and multi-frame estimation, the dilated temporal model is employed to process varying skeleton sequences. Also, importantly, we carefully design the interleaving of spatial semantics with temporal dependencies to achieve a synergistic effect. To this end, we propose a simple yet effective graph attention spatio-temporal convolutional network (GAST-Net) that comprises of interleaved temporal convolutional and graph attention blocks. Experiments on two challenging benchmark datasets (Human3.6M and HumanEva-I) and YouTube videos demonstrate that our approach effectively mitigates depth ambiguity and self-occlusion, generalizes to half upper body estimation, and achieves competitive performance on 2D-to-3D video pose estimation. Code, video, and supplementary information is available at: \href{http://www.juanrojas.net/gast/}{http://www.juanrojas.net/gast/}
△ Less
Submitted 19 October, 2020; v1 submitted 11 March, 2020;
originally announced March 2020.