arXiv subjects

Jin Yang

Publications and source records attributed to Jin Yang.

At least 109 records · Page 6Linked to original sources

arXiv2021

A Highly Configurable Hardware/Software Stack for DNN Inference Acceleration

This work focuses on an efficient Agile design methodology for domain-specific accelerators. We employ feature-by-feature enhancement of a vertical development stack and apply it to the TVM/VTA inference accelerator. We have enhanced the VTA design space and enabled end-to-end support for additional workloads. This has been accomplished by augmenting the VTA micro-architecture and instruction set architecture (ISA), as well as by enhancing the TVM compilation stack to support a wide range of VTA configs. The VTA tsim implementation (CHISEL-based) has been enhanced with fully pipelined versions of the ALU/GEMM execution units. In tsim, memory width can now range between 8-64 bytes. Field widths have been made more flexible to support larger scratchpads. New instructions have been added: element-wise 8-bit multiplication to support depthwise convolution, and load with a choice of pad values to support max pooling. Support for more layers and better double buffering has also been added. Fully pipelining ALU/GEMM helps significantly: 4.9x fewer cycles with minimal area change to run ResNet-18 under the default config. Configs featuring a further 11.5x decrease in cycle count at a cost of 12x greater area can be instantiated. Many points on the area-performance pareto curve are shown, showcasing the balance of execution unit sizing, memory interface width, and scratchpad sizing. Finally, VTA is now able to run Mobilenet 1.0 and all layers for ResNets, including the previously disabled pooling and fully connected layers. The TVM/VTA architecture has always featured end-to-end workload evaluation on RTL in minutes. With our modifications, it now offers a much greater number of feasible configurations with a wide range of cost vs. performance. All capabilities mentioned are available in opensource forks while a subset of these capabilities have already been upstreamed.

cs.AR↗

arXiv2021

Ground states of 2D tilted dipolar bosons with density-induced hopping

Motivated by recent experiments with ultracold magnetic atoms trapped in optical lattices where the orientation of atomic dipoles can be fully controlled by external fields, we study the ground state properties of dipolar bosons trapped in a two-dimensional lattice with density-induced hopping and where the dipoles are tilted along the $xz$ plane. We present ground state phase diagrams of the above system at different tilt angles. We find that, as the dipolar interaction increases, the superfluid phase at half filling factor is destroyed in favor of either a checkerboard or stripe solid phase for tilt angle $θ\lesssim 35.3^{\circ}$ or $θ\gtrsim 35.3^{\circ}$ respectively. More interesting physics happens at tilt angles $θ\gtrsim 58^{\circ}$, where we find that, as the dipolar interaction strength increases, solid phases first appear at filling factor lower than $0.5$. Moreover, unlike what observed at lower tilt angles, we find that, at half filling, a stripe supersolid intervenes between the superfluid and stripe solid phase.

cond-mat.quant-gas↗

arXiv2021

Two-stage short-term wind power forecasting algorithm using different feature-learning models

Two-stage ensemble-based forecasting methods have been studied extensively in the wind power forecasting field. However, deep learning-based wind power forecasting studies have not investigated two aspects. In the first stage, different learning structures considering multiple inputs and multiple outputs have not been discussed. In the second stage, the model extrapolation issue has not been investigated. Therefore, we develop four deep neural networks for the first stage to learn data features considering the input-and-output structure. We then explore the model extrapolation issue in the second stage using different modeling methods. Considering the overfitting issue, we propose a new moving window-based algorithm using a validation set in the first stage to update the training data in both stages with two different moving window processes.Experiments were conducted at three wind farms, and the results demonstrate that the model with single input multiple output structure obtains better forecasting accuracy compared to existing models. In addition, the ridge regression method results in a better ensemble model that can further improve forecasting accuracy compared to existing machine learning methods. Finally, the proposed two-stage forecasting algorithm can generate more accurate and stable results than existing algorithms.

eess.SP↗

arXiv2021

Deep Reinforcement Learning with a Stage Incentive Mechanism of Dense Reward for Robotic Trajectory Planning

(This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.) To improve the efficiency of deep reinforcement learning (DRL)-based methods for robot manipulator trajectory planning in random working environments, we present three dense reward functions. These rewards differ from the traditional sparse reward. First, a posture reward function is proposed to speed up the learning process with a more reasonable trajectory by modeling the distance and direction constraints, which can reduce the blindness of exploration. Second, a stride reward function is proposed to improve the stability of the learning process by modeling the distance and movement distance of joint constraints. Finally, in order to further improve learning efficiency, we are inspired by the cognitive process of human behavior and propose a stage incentive mechanism, including a hard stage incentive reward function and a soft stage incentive reward function. Extensive experiments show that the soft stage incentive reward function is able to improve the convergence rate by up to 46.9% with the state-of-the-art DRL methods. The percentage increase in the convergence mean reward was 4.4-15.5% and the percentage decreases with respect to standard deviation were 21.9-63.2%. In the evaluation experiments, the success rate of trajectory planning for a robot manipulator reached 99.6%.

cs.AI↗

arXiv2021

Predicting Complex Non-spherical Instability Shapes of Inertial Cavitation Bubbles in Viscoelastic Soft Matter

Inertial cavitation in soft matter is an important phenomenon featured in a wide array of biological and engineering processes. Recent advances in experimental, theoretical, and numerical techniques have provided access into a world full of nonlinear physics, yet most of our quantitative understanding to date has been centered on a spherically symmetric description of the cavitation process. However, cavitation bubble growth and collapse rarely occur in a perfectly symmetrical fashion, particularly in soft materials. Predicting the onset of dynamically arising, non-spherical instabilities has remained a significant, unresolved challenge in part due to the additional constitutive complexities introduced by the surrounding nonlinear viscoelastic solid. Here, we provide a new theoretical model capable of accurately predicting the onset of non-spherical instability shapes of a bubble in a soft material by explicitly accounting for all pertinent nonlinear interactions between the fluid-like cavitation bubble and the solid-like surroundings. Comparison against high-resolution experimental images from laser-induced cavitation events in a polyacrylamide (PA) hydrogel show excellent agreement. Interestingly, and consistent with experimental findings, our model predicts the emergence of various dynamic instability shapes for hoop stretch ratios greater than one in contrast to most quasi-static investigations. Our new theoretical framework not only provides unprecedented insight into the cavitation dynamics in a soft solid, but it also provides a quantitative means of interpreting bubble dynamics relevant to a wide array of engineering and medical applications as well as natural phenomena.

cond-mat.soft↗

arXiv2021

Robust wavefront dislocations of Friedel oscillations in gapped graphene

Friedel oscillation is a well-known wave phenomenon, which represents the oscillatory response of electron waves to imperfection. By utilizing the pseudospin-momentum locking in gapless graphene, two recent experiments demonstrate the measurement of the topological Berry phase by corresponding to the unique number of wavefront dislocations in Friedel oscillations. Here, we study the Friedel oscillations in gapped graphene, in which the pseudospin-momentum locking is broken. Unusually, the wavefront dislocations do occur as that in gapless graphene, which expects the immediate verification in the current experimental condition. The number of wavefront dislocations is ascribed to the invariant pseudospin winding number in gaped and gapless graphene. This study deepens the understanding of correspondence between topological quantity and wavefront dislocations in Friedel oscillations, and implies the possibility to observe the wavefront dislocations of Friedel oscillations in intrinsic gapped two-dimensional materials, e.g., transition metal dichalcogenides.

cond-mat.mes-hall↗

arXiv2021

Site-resolved imaging of ultracold fermions in a triangular-lattice quantum gas microscope

Quantum gas microscopes have expanded the capabilities of quantum simulation of Hubbard models by enabling the study of spatial spin and density correlations in square lattices. However, quantum gas microscopes have not been realized for fermionic atoms in frustrated geometries. Here, we demonstrate the single-atom resolved imaging of ultracold fermionic $^{6}$Li atoms in a triangular optical lattice with a lattice constant of 1003 nm. The optical lattice is formed by a recycled narrow-linewidth, high-power laser combined with a light sheet to allow for Raman sideband cooling on the $D_1$ line. We optically resolve single atoms on individual lattice sites using a high-resolution objective to collect scattered photons while cooling them close to the two-dimensional ground vibrational level in each lattice site. By reconstructing the lattice occupation, we measure an imaging fidelity of ~98%. Our new triangular lattice microscope platform for fermions clears the path for studying spin-spin correlations, entanglement and dynamics of geometrically frustrated Hubbard systems which are expected to exhibit exotic emergent phenomena including spin liquids and kinetic frustration.

cond-mat.quant-gas↗

arXiv2021

Model Synthesis for Communication Traces of System-on-Chip Designs

Concise and abstract models of system-level behaviors are invaluable in design analysis, testing, and validation. In this paper, we consider the problem of inferring models from communication traces of system-on-chip~(SoC) designs. The traces capture communications among different blocks of a SoC design in terms of messages exchanged. The extracted models characterize the system-level communication protocols governing how blocks exchange messages, and coordinate with each other to realize various system functions. In this paper, the above problem is formulated as a constraint satisfaction problem, which is then fed to a SMT solver. The solutions returned by the SMT solver are used to extract the models that accept the input traces. In the experiments, we demonstrate the proposed approach with traces collected from a transaction-level simulation model of a multicore SoC design and traces of a more detailed multicore SoC design developed in GEM5 environment.

cs.IR↗

arXiv2020

Acoustic Cavitation Rheometry

Characterization of soft materials is challenging due to their high compliance and the strain-rate dependence of their mechanical properties. The inertial microcavitation-based high strain-rate rheometry (IMR) method [Estrada et al., J. Mech. Phys. Solids, 2018, 112, 291-317] combines laser-induced cavitation measurements with a model for the bubble dynamics to measure local properties of polyacrylamide hydrogel under high strain-rates from $10^3$ to $10^8$ s$^{-1}$. While promising, laser-induced cavitation involves plasma formation and optical breakdown during nucleation, a process that could alter local material properties before measurements are obtained. In the present study, we extend the IMR method to another means to generate cavitation, namely high-amplitude focused ultrasound, and apply the resulting acoustic-cavitation-based IMR to characterize the mechanical properties of agarose hydrogels. Material properties including viscosity, elastic constants, and a stress-free bubble radius are inferred from bubble radius histories in $0.3\%$ and $1\%$ agarose gels. An ensemble-based data assimilation is used to further help interpret the obtained estimates. The resulting parameter distributions are consistent with available measurements of agarose gel properties and with expected trends related to gel concentration and high strain-rate loading. Our findings demonstrate the utility of applying IMR and data assimilation methods with single-bubble acoustic cavitation data for measurement of viscoelastic properties.

cond-mat.soft↗

arXiv2020

Reconfigurable Intelligent Surface: Design the Channel -- a New Opportunity for Future Wireless Networks

In this paper, we survey state-of-the-art research outcomes in the burgeoning field of reconfigurable intelligent surface (RIS) in view of its potential for significant performance enhancement for next generation wireless communication networks by means of adapting the propagation environment. Emphasis has been placed on several aspects gating the commercially viability of a future network deployment. Comprehensive summaries are provided for practical hardware design considerations and broad implications of artificial intelligence techniques, so are in-depth outlooks on salient aspects of system models, use cases, and physical layer optimization techniques.

eess.SP↗

arXiv2020

A Digital Twin for Reconfigurable Intelligent Surface Assisted Wireless Communication

Reconfigurable Intelligent Surface (RIS) has emerged as one of the key technologies for 6G in recent years, which comprise a large number of low-cost passive elements that can smartly interact with the impinging electromagnetic waves for performance enhancement. However, optimally configuring massive number of RIS elements remains a challenge. In this paper, we present a novel digital-twin framework for RIS-assisted wireless networks which we name it Environment-Twin (Env-Twin). The goal of the Env-Twin framework is to enable automation of optimal control at various granularities. In this paper, we present one example of the Env-Twin models to learn the mapping function between the RIS configuration with measured attributes for the receiver location, and the corresponding achievable rate in an RIS-assisted wireless network without involving explicit channel estimation or beam training overhead. Once learned, our Env-Twin model can be used to predict optimal RIS configuration for any new receiver locations in the same wireless network. We leveraged deep learning (DL) techniques to build our model and studied its performance and robustness. Simulation results demonstrate that the proposed Env-Twin model can recommend near-optimal RIS configurations for test receiver locations which achieved close to an upper bound performance that assumes perfect channel knowledge. Our Env-Twin model was trained using less than 2% of the total receiver locations. This promising result represents great potential of the proposed Env-Twin framework for developing a practical RIS solution where the panel can automatically configure itself without requesting channel state information (CSI) from the wireless network infrastructure.

cs.NI↗

arXiv2020

Characterizing viscoelastic materials via ensemble-based data assimilation of bubble collapse observations

Viscoelastic material properties at high strain rates are needed to model many biological and medical systems. Bubble cavitation can induce such strain rates, and the resulting bubble dynamics are sensitive to the material properties. Thus, in principle, these properties can be inferred via measurements of the bubble dynamics. Estrada et al. (2018) demonstrated such bubble-dynamic high-strain-rate rheometry by using least-squares shooting to minimize the difference between simulated and experimental bubble radius histories. We generalize their technique to account for additional uncertainties in the model, initial conditions, and material properties needed to uniquely simulate the bubble dynamics. Ensemble-based data assimilation minimizes the computational expense associated with the bubble cavitation model. We test an ensemble Kalman filter (EnKF), an iterative ensemble Kalman smoother (IEnKS), and a hybrid ensemble-based 4D--Var method (En4D--Var) on synthetic data, assessing their estimations of the viscosity and shear modulus of a Kelvin--Voigt material. Results show that En4D--Var and IEnKS provide better moduli estimates than EnKF. Applying these methods to the experimental data of Estrada et al. (2018) yields similar material property estimates to those they obtained, but provides additional information about uncertainties. In particular, the En4D--Var yields lower viscosity estimates for some experiments, and the dynamic estimators reveal a potential mechanism that is unaccounted for in the model, whereby the viscosity is reduced in some cases due to material damage occurring at bubble collapse.

physics.comp-ph↗

arXiv2020

An Origin of Dzyaloshinskii-Moriya Interaction at Graphene-Ferromagnet Interfaces Due to the Intralayer RKKY/BR Interaction

We present a theory of both the itinerant carrier-mediated RKKY interaction and the virtual excitations-mediated Bloembergen-Rowland (BR) interaction between magnetic moments in graphene induced by proximity effect with a ferromagnetic film. We show that the RKKY/BR interaction consists of the Heisenberg, Ising, and Dzyaloshinskii-Moriya (DM) terms. In the case of the nearest distance, we estimate the DM term from the RKKY/BR interaction is about 0.13 meV for the graphene/Co interface, which is consistent with the experimental result of DM interaction $0.16 \pm 0.05$ meV. Our calculations indicate that the intralayer RKKY/BR interaction may be a possible physical origin of the DM interaction in the graphene-ferromagnet interface. This work provides a new perspective to comprehend the DM interaction in graphene/ferromagnet systems.

cond-mat.mes-hall↗

arXiv2020

Arbitrary manipulations of dual harmonics and their wave behaviors based on space-time-coding digital metasurface

Space-time modulated metasurfaces have attracted significant attention due to the additional degree of freedom in manipulating the electromagnetic (EM) waves in both space and time domains. However, the existing techniques have limited wave control capabilities, leading to just a few feasible schemes like regulation of only one specific harmonic. Here, we propose to realize independent manipulations of arbitrarily dual harmonics and their wave behaviors using a space-time-coding (STC) digital metasurface. By employing different STC sequences to the reflection phase of the metasurface, independent phase-pattern configurations of two desired harmonics can be achieved simultaneously, which further leads to independent beam shaping at the two harmonic frequencies. An analytical theory is developed to offer the physical insights in the arbitrary dual-harmonic manipulations of spectra and spatial beams, which is verified by experiments with good agreements. The presented STC strategy provides a new way to design multifunctional programmable systems, which will find potential applications such as cognitive radar and multi-user wireless communications.

physics.app-ph↗

arXiv2020

Mining Message Flows from System-on-Chip Execution Traces

Comprehensive and well-defined specifications are necessary to perform rigorous and thorough validation of system-on-chip (SoC) designs. Message flows specify how components of an SoC design communicate and coordinate with each other to realize various system functions. Message flow specifications are essential for efficient system-level validation and debug for SoC designs. However, in practice such specifications are usually not available, often ambiguous, incomplete, or even contain errors. This paper addresses that problem by proposing a specification mining framework, FlowMiner, that automatically extracts message flows from SoC execution traces, which, unlike software traces, show a high degree of concurrency. A set of inference rules and optimization techniques are presented to improve mining performance and reduce mining complexity. Evaluation of this framework in several experiments shows promising results.

cs.SE↗

arXiv2020

A Post-Silicon Trace Analysis Approach for System-on-Chip Protocol Debug

Reconstructing system-level behavior from silicon traces is a critical problem in post-silicon validation of System-on-Chip designs. Current industrial practice in this area is primarily manual, depending on collaborative insights of the architects, designers, and validators. This paper presents a trace analysis approach that exploits architectural models of the system-level protocols to reconstruct design behavior from partially observed silicon traces in the presence of ambiguous and noisy data. The output of the approach is a set of all potential interpretations of a system's internal executions abstracted to the system-level protocols. To support the trace analysis approach, a companion trace signal selection framework guided by system-level protocols is also presented, and its impacts on the complexity and accuracy of the analysis approach are discussed. That approach and the framework have been evaluated on a multi-core system-on-chip prototype that implements a set of common industrial system-level protocols.

cs.AR↗

arXiv2020

Mining Message Flows using Recurrent Neural Networks for System-on-Chip Designs

Comprehensive specifications are essential for various activities across the entire validation continuum for system-on-chip (SoC) designs. However, specifications are often ambiguous, incomplete, or even contain inconsistencies or errors. This paper addresses this problem by developing a specification mining approach that automatically extracts sequential patterns from SoC transaction-level traces such that the mined patterns collectively characterize system-level specifications for SoC designs. This approach exploits long short-term memory (LSTM) networks trained with the collected SoC execution traces to capture sequential dependencies among various communication events. Then, a novel algorithm is developed to efficiently extract sequential patterns on system-level communications from the trained LSTM models. Several trace processing techniques are also proposed to enhance the mining performance. We evaluate the proposed approach on simulation traces of a non-trivial multi-core SoC prototype. Initial results show that the proposed approach is capable of extracting various patterns on system-level specifications from the highly concurrent SoC execution traces.

cs.DC↗

arXiv2020

Single upper limb pose estimation method based on improved stacked hourglass network

At present, most high-accuracy single-person pose estimation methods have high computational complexity and insufficient real-time performance due to the complex structure of the network model. However, a single-person pose estimation method with high real-time performance also needs to improve its accuracy due to the simple structure of the network model. It is currently difficult to achieve both high accuracy and real-time performance in single-person pose estimation. For use in human-machine cooperative operations, this paper proposes a single-person upper limb pose estimation method based on an end-to-end approach for accurate and real-time limb pose estimation. Using the stacked hourglass network model, a single-person upper limb skeleton key point detection model was designed.Deconvolution was employed to replace the up-sampling operation of the hourglass module in the original model, solving the problem of rough feature maps. Integral regression was used to calculate the position coordinates of key points of the skeleton, reducing quantization errors and calculations. Experiments showed that the developed single-person upper limb skeleton key point detection model achieves high accuracy and that the pose estimation method based on the end-to-end approach provides high accuracy and real-time performance.

cs.CV↗