arXiv subjects

Mingyu Wang

Publications and source records attributed to Mingyu Wang.

27 records · Page 2Linked to original sources

arXiv2022

Game-Theoretic Planning for Autonomous Driving among Risk-Aware Human Drivers

We present a novel approach for risk-aware planning with human agents in multi-agent traffic scenarios. Our approach takes into account the wide range of human driver behaviors on the road, from aggressive maneuvers like speeding and overtaking, to conservative traits like driving slowly and conforming to the right-most lane. In our approach, we learn a mapping from a data-driven human driver behavior model called the CMetric to a driver's entropic risk preference. We then use the derived risk preference within a game-theoretic risk-sensitive planner to model risk-aware interactions among human drivers and an autonomous vehicle in various traffic scenarios. We demonstrate our method in a merging scenario, where our results show that the final trajectories obtained from the risk-aware planner generate desirable emergent behaviors. Particularly, our planner recognizes aggressive human drivers and yields to them while maintaining a greater distance from them. In a user study, participants were able to distinguish between aggressive and conservative simulated drivers based on trajectories generated from our risk-sensitive planner. We also observe that aggressive human driving results in more frequent lane-changing in the planner. Finally, we compare the performance of our modified risk-aware planner with existing methods and show that modeling human driver behavior leads to safer navigation.

cs.RO↗

arXiv2021

Maximum-Entropy Multi-Agent Dynamic Games: Forward and Inverse Solutions

In this paper, we study the problem of multiple stochastic agents interacting in a dynamic game scenario with continuous state and action spaces. We define a new notion of stochastic Nash equilibrium for boundedly rational agents, which we call the Entropic Cost Equilibrium (ECE). We show that ECE is a natural extension to multiple agents of Maximum Entropy optimality for single agents. We solve both the "forward" and "inverse" problems for the multi-agent ECE game. For the forward problem, we provide a Riccati algorithm to compute closed-form ECE feedback policies for the agents, which are exact in the Linear-Quadratic-Gaussian case. We give an iterative variant to find locally ECE feedback policies for the nonlinear case. For the inverse problem, we present an algorithm to infer the cost functions of the multiple interacting agents given noisy, boundedly rational input and state trajectory examples from agents acting in an ECE. The effectiveness of our algorithms is demonstrated in a simulated multi-agent collision avoidance scenario, and with data from the INTERACTION traffic dataset. In both cases, we show that, by taking into account the agents' game theoretic interactions using our algorithm, a more accurate model of agents' costs can be learned, compared with standard inverse optimal control methods.

math.OC↗

arXiv2021

SuperVoxHenry Tucker-Enhanced and FFT-Accelerated Inductance Extraction for Voxelized Superconducting Structures

This paper introduces SuperVoxHenry, an inductance extraction simulator for analyzing voxelized superconducting structures. SuperVoxHenry extends the capabilities of the inductance extractor VoxHenry for analyzing the superconducting structures by incorporating the following enhancements. 1. SuperVoxHenry utilizes a two-fluid model to account for normal currents and supercurrents. 2. SuperVoxHenry introduces the Tucker decompositions to reduce the memory requirement of circulant tensors as well as the setup time of the simulator. 3. SuperVoxHenry incorporates an aggregation-based algebraic multigrid technique to obtain the sparse preconditioner.

cs.CE↗

arXiv2021

Compression of Far-Fields in the Fast Multipole Method via Tucker Decomposition

Tucker decomposition is proposed to reduce the memory requirement of the far-fields in the fast multipole method (FMM)-accelerated surface integral equation simulators. It is particularly used to compress the far-fields of FMM groups, which are stored in three-dimensional (3-D) arrays (or tensors). The compressed tensors are then used to perform fast tensor-vector multiplications during the aggregation and disaggregation stages of the FMM. For many practical scenarios, the proposed Tucker decomposition yields a significant reduction in the far-fields' memory requirement while dramatically accelerating the aggregation and disaggregation stages. For the electromagnetic scattering analysis of a 30λ-diameter sphere, it reduces the memory requirement of the far-fields more than 87% while it expedites the aggregation and disaggregation stages by a factor of 15.8 and 15.2, respectively, where λ is the wavelength in free space.

physics.comp-ph↗

arXiv2020

VoxCap: FFT-Accelerated and Tucker-Enhanced Capacitance Extraction Simulator for Voxelized Structures

VoxCap, a fast Fourier transform (FFT)-accelerated and Tucker-enhanced integral equation simulator for capacitance extraction of voxelized structures, is proposed. The VoxCap solves the surface integral equations (SIEs) for conductor and dielectric surfaces with three key attributes that make the VoxCap highly CPU and memory efficient for the capacitance extraction of the voxelized structures: (i) VoxCap exploits the FFTs for accelerating the matrix-vector multiplications during the iterative solution of linear system of equations arising due to the discretization of SIEs. (ii) During the iterative solution, VoxCap uses a highly effective and memory-efficient preconditioner that reduces the number of iterations significantly. (iii) VoxCap employs Tucker decompositions to compress the block Toeplitz and circulant tensors, requiring the largest memory in the simulator. By doing so, it reduces the memory requirement of these tensors from hundreds of gigabytes to a few megabytes and the CPU time required to obtain Toeplitz tensors from tens of minutes (even hours) to a few seconds for very large scale problems. VoxCap is capable of accurately computing capacitance of arbitrarily shaped and large-scale voxelized structures on a desktop computer.

cs.CE↗

arXiv2020

Directional Primitives for Uncertainty-Aware Motion Estimation in Urban Environments

We can use driving data collected over a long period of time to extract rich information about how vehicles behave in different areas of the roads. In this paper, we introduce the concept of directional primitives, which is a representation of prior information of road networks. Specifically, we represent the uncertainty of directions using a mixture of von Mises distributions and associated speeds using gamma distributions. These location-dependent primitives can be combined with motion information of surrounding vehicles to predict their future behavior in the form of probability distributions. Experiments conducted on highways, intersections, and roundabouts in the Carla simulator, as well as real-world urban driving datasets, indicate that primitives lead to better uncertainty-aware motion estimation.

cs.RO↗

arXiv2020

Phoenix: A Low-Precision Floating-Point Quantization Oriented Architecture for Convolutional Neural Networks

Convolutional neural networks (CNNs) achieve state-of-the-art performance at the cost of becoming deeper and larger. Although quantization (both fixed-point and floating-point) has proven effective for reducing storage and memory access, two challenges -- 1) accuracy loss caused by quantization without calibration, fine-tuning or re-training for deep CNNs and 2) hardware inefficiency caused by floating-point quantization -- prevent processors from completely leveraging the benefits. In this paper, we propose a low-precision floating-point quantization oriented processor, named Phoenix, to address the above challenges. We primarily have three key observations: 1) 8-bit floating-point quantization incurs less error than 8-bit fixed-point quantization; 2) without using any calibration, fine-tuning or re-training techniques, normalization before quantization further reduces accuracy degradation; 3) 8-bit floating-point multiplier achieves higher hardware efficiency than 8-bit fixed-point multiplier if the full-precision product is applied. Based on these key observations, we propose a normalization-oriented 8-bit floating-point quantization method to reduce storage and memory access with negligible accuracy loss (within 0.5%/0.3% for top-1/top-5 accuracy, respectively). We further design a hardware processor to address the hardware inefficiency caused by floating-point multiplier. Compared with a state-of-the-art accelerator, Phoenix is 3.32x and 7.45x better in performance with the same core area for AlexNet and VGG16, respectively.

eess.SP↗

arXiv2020

Low Precision Floating-point Arithmetic for High Performance FPGA-based CNN Acceleration

Low precision data representation is important to reduce storage size and memory access for convolutional neural networks (CNNs). Yet, existing methods have two major limitations: (1) requiring re-training to maintain accuracy for deep CNNs, and (2) needing 16-bit floating-point or 8-bit fixed-point for a good accuracy. In this paper, we propose a low precision (8-bit) floating-point (LPFP) quantization method for FPGA-based acceleration to overcome the above limitations. Without any re-training, LPFP finds an optimal 8-bit data representation with negligible top-1/top-5 accuracy loss (within 0.5%/0.3% in our experiments, respectively, and significantly better than existing methods for deep CNNs). Furthermore, we implement one 8-bit LPFP multiplication by one 4-bit multiply-adder (MAC) and one 3-bit adder, and therefore implement four 8-bit LPFP multiplications using one DSP slice of Xilinx Kintex 7 family (KC705 in this paper) while one DSP can implement only two 8-bit fixed-point multiplications. Experiments on six typical CNNs for inference show that on average, we improve throughput by 64.5x over Intel i9 CPU and by 1.5x over existing FPGA accelerators. Particularly for VGG16 and YOLO, compared to six recent FPGA accelerators, we improve average throughput by 3.5x and 27.5x and improve average throughput per DSP by 4.1x and 5x, respectively. To the best of our knowledge, this is the first in-depth study to simplify one multiplication for CNN inference to one 4-bit MAC and implement four multiplications within one DSP while maintaining comparable accuracy without any re-training.

eess.SP↗

arXiv2017

Stepwise quantized surface states and delayed Landau level hybridization in Co cluster-decorated BiSbTeSe2 topological insulator devices

In three-dimensional topological insulators (TIs), the nontrivial topology in their electronic bands casts a gapless state on their solid surfaces, using which dissipationless TI edge devices based on the quantum anomalous Hall (QAH) effect and quantum Hall (QH) effect have been demonstrated. Practical TI devices present a pair of parallel-transport topological surface states (TSSs) on their top and bottom surfaces. However, due to the no-go theorem, the two TSSs always appear as a pair and are expected to quantize synchronously. Quantized transport of a separate Dirac channel is still desirable, but has never been observed in graphene even after intense investigation over a period of 13 years, with the potential aim of half-QHE. By depositing Co atomic clusters, we achieved stepwise quantization of the top and bottom surfaces in BiSbTeSe2 (BSTS) TI devices. Renormalization group flow diagrams13, 22 (RGFDs) reveal two sets of converging points (CVPs) in the (Gxy, Gxx) space, where the top surface travels along an anomalous quantization trajectory while the bottom surface retains 1/2 e2/h. This results from delayed Landau-level (LL) hybridization (DLLH) due to coupling between Co clusters and TSS Fermions.

cond-mat.mes-hall↗