arXiv subjects

Yuan Meng

Publications and source records attributed to Yuan Meng.

86 records · Page 5Linked to original sources

arXiv2022

Intelligent optoelectronic processor for orbital angular momentum spectrum measurement

Orbital angular momentum (OAM) detection underpins almost all aspects of vortex beams' advances such as communication and quantum analogy. Conventional schemes are frustrated by low speed, complicated system, limited detection range. Here, we devise an intelligent processor composed of photonic and electronic neurons for OAM spectrum measurement in a fast, accurate and direct manner. Specifically, optical layers extract invisible topological charge information from incoming light and a shallow electronic layer predicts the exact spectrum. The integration of optical-computing promises us a compact single-shot system with high speed and energy efficiency, neither necessitating reference wave nor repetitive steps. Importantly, our processor is endowed with salient generalization ability and robustness against diverse structured light and adverse effects. We further raise a universal model interpretation paradigm to reveal the underlying physical mechanisms in the hybrid processor, as distinct from conventional 'black-box' networks. Such interpretation algorithm can improve the detection efficiency. We also complete the theory of optoelectronic network enabling its efficient training. This work not only contributes to the explorations on OAM physics and applications, and also broadly inspires the advanced links between intelligent computing and physical effects.

cs.ET↗

arXiv2022

Accelerating Monte-Carlo Tree Search on CPU-FPGA Heterogeneous Platform

Monte Carlo Tree Search (MCTS) methods have achieved great success in many Artificial Intelligence (AI) benchmarks. The in-tree operations become a critical performance bottleneck in realizing parallel MCTS on CPUs. In this work, we develop a scalable CPU-FPGA system for Tree-Parallel MCTS. We propose a novel decomposition and mapping of MCTS data structure and computation onto CPU and FPGA to reduce communication and coordination. High scalability of our system is achieved by encapsulating in-tree operations in an SRAM-based FPGA accelerator. To lower the high data access latency and inter-worker synchronization overheads, we develop several hardware optimizations. We show that by using our accelerator, we obtain up to $35\times$ speedup for in-tree operations, and $3\times$ higher overall system throughput. Our CPU-FPGA system also achieves superior scalability wrt number of parallel workers than state-of-the-art parallel MCTS implementations on CPU.

cs.DC↗

arXiv2022

An Empirical Study of Pseudo-Labeling for Image-based 3D Object Detection

Image-based 3D detection is an indispensable component of the perception system for autonomous driving. However, it still suffers from the unsatisfying performance, one of the main reasons for which is the limited training data. Unfortunately, annotating the objects in the 3D space is extremely time/resource-consuming, which makes it hard to extend the training set arbitrarily. In this work, we focus on the semi-supervised manner and explore the feasibility of a cheaper alternative, i.e. pseudo-labeling, to leverage the unlabeled data. For this purpose, we conduct extensive experiments to investigate whether the pseudo-labels can provide effective supervision for the baseline models under varying settings. The experimental results not only demonstrate the effectiveness of the pseudo-labeling mechanism for image-based 3D detection (e.g. under monocular setting, we achieve 20.23 AP for moderate level on the KITTI-3D testing set without bells and whistles, improving the baseline model by 6.03 AP), but also show several interesting and noteworthy findings (e.g. the models trained with pseudo-labels perform better than that trained with ground-truth annotations based on the same training data). We hope this work can provide insights for the image-based 3D detection community under a semi-supervised setting. The codes, pseudo-labels, and pre-trained models will be publicly available.

cs.CV↗

arXiv2022

Photon regions, shadow observables and constraints from M87* of a charged rotating black hole

Inspired by the observations of supermassive black hole M87* in \emph{Event Horizon Telescope }(EHT) experiment, a remarkable surge in black hole physics is to use the black hole shadow's observables to distinguish general relativity (GR) and modified theories of gravity (MoG), which could also help to disclose the astrophysical nature of the center black hole in EHT observation. In this paper, we shall extensively carry out the study of a charged rotating black hole in conformal gravity, in which the term related with the charge has different falloffs from the usual Kerr-Newman (KN) black hole. We investigate the spacetime properties including the horizons, ergospheres and the photon regions; afterward, we show the boundary of black hole shadow and investigate its characterized observables. The features closely depend on the spin and charge parameters, which are compared with those in Kerr and KN black holes. Then presupposing the M87* a charged rotating black hole in conformal gravity, we also constrain the black hole parameters via the observation constraints from EHT experiment. We find that the constraints on the inferred circularity deviation, $ΔC \lesssim 0.1$, and on the shadow axial ratio, $1< D_x \lesssim 4/3$, for the M87* black hole are satisfied for the entire parameter space of the charged rotating black hole in conformal gravity. However, the shadow angular diameter $θ_d = 42 \pm 3 μas$ will give upper bound on the parameter space. Our findings indicate that the current charged rotating black hole in conformal gravity could be a candidate for astrophysical black holes. Moreover, the EHT observation on the axial ratio $D_x$ may help us to distinguish Kerr black hole and the current charged rotating black hole in conformal gravity in some parameter space.

gr-qc↗

arXiv2021

Guided mode meta-optics: Metasurface-dressed nanophotonic waveguides for arbitrary designer mode couplers and on-chip OAM emitters with configurable topological charge

Metasurfaces have achieved fruitful results in tailoring complexing light fields in free space. However, a systematic investigation on applying the concept of meta-optics to completely control waveguide modes is still elusive. Here we present a comprehensive catalog capable of selectively and exclusively excite almost arbitrary high-order waveguide modes of interest, leveraging silicon metasurface-patterned silicon nitride waveguides. By simultaneously engineering the phase-matched gradient of the metasurface and the vectorial spatial modal overlap between the nanoantenna near-field and target waveguide mode for excitation, either single or multiple high-order modes are successfully launched with high purity reaching 98% and broad bandwidth. Moreover, on-chip twisted light generators are also theoretically demonstrated with configurable OAM topological charge \ell from -3 to +2, serving as a comprehensive framework for metasurface-enabled guided mode optics and motivating further applications such as versatile integrated couplers, demultiplexers, and mode-division multiplexing-based communication systems.

physics.optics↗

arXiv2021

Deep-learning-based recognition of multi-singularity structured light

Structured light with customized complex topological pattern inspires diverse classical and quantum investigations underpinned by accurate detection techniques. However, the current detection schemes are limited to vortex beam with simple phase singularity. The precise recognition of general structured light with multiple singularities remains elusive. Here, we report a deep learning (DL) framework that can unveil multi-singularity phase structures in an end-to-end manner after feeding only two intensity patterns upon beam propagation captured via a camera, thus unleashing intuitive information of twisted photons. The DL toolbox can also acquire phases of Laguerre-Gaussian (LG) modes with single singularity and other general phase objects likewise. Leveraging this DL platform, a phase-based optical secret sharing (OSS) protocol is proposed, which is based on a more general class of multi-singularity modes than conventional LG beams. The OSS protocol features strong security, wealthy state space and convenient intensity-based measurements. This study opens new avenues for vortex beam communications, laser mode analysis, microscopy, Bose-Einstein condensates characterization, etc.

physics.optics↗

arXiv2021

DYNAMAP: Dynamic Algorithm Mapping Framework for Low Latency CNN Inference

Most of the existing work on FPGA acceleration of Convolutional Neural Network (CNN) focus on employing a single strategy (algorithm, dataflow, etc.) across all the layers. Such an approach does not achieve optimal latency on complex and deep CNNs. Emerging CNNs have diverse per-layer computation characteristics including parallelism, arithmetic intensity, locality, and memory footprint. Per-layer strategy selection and fine-grained tuning are required to achieve low end-to-end latency. However, specialized hardware modules dedicated to each layer limit the per-layer utilization and adversely affect end-to-end latency. In this paper, we address these problems by an algorithm-architecture co-optimization framework, DYNAMAP, consisting of (1) a unified hardware overlay that can be reused across layers, supporting dynamic mapping of all three families of popular convolution algorithms, and further allowing flexible dataflow switching to maximize hardware utilization for each layer; (2) a novel software Design Space Exploration (DSE) flow that customizes the hardware overlay and chooses optimal strategy mapping. We show that the algorithm mapping space increases exponentially with network depth, and while the optimal algorithm selection problem is NP-hard in general, by exploiting the series-parallel structure of CNN models, we demonstrate a polynomial-time solution for optimal algorithm mapping. DYNAMAP is optimized for any CNN, including those having diverse computation and memory requirements across the layers. We demonstrate DYNAMAP using two state-of-the-art CNNs - GoogleNet and Inception-V4. The generated accelerators achieve up to $2.8\times$ and $1.4\times$ speedups, respectively, wrt inference latency compared with the state-of-the-art FPGA implementations.

cs.DC↗

arXiv2020

Ultrafast two-photon emission in a doped semiconductor thin film

As a high-order quantum transition, two-photon emission has an extremely low occurrence rate compared to one-photon emission, thus having been considered a forbidden process. Here, we propose a scheme that allows ultrafast two-photon emission, leveraging highly confined surface plasmon polariton modes in a degenerately-doped, light-emitting semiconductor thin film. The surface plasmon polariton modes are tailored to have simultaneous spectral and spatial overlap with the two-photon emission in the semiconductor. Using degenerately-doped InSb as the prototype material, we show that the two-photon emission can be accelerated by 10 orders of magnitude: from tens of milliseconds to picoseconds, surpassing the one-photon emission rate. Our result provides a semiconductor platform for ultrafast single and entangled photon generation, with a tunable emission wavelength in the mid-infrared.

physics.optics↗

arXiv2019

Estimating Granger Causality with Unobserved Confounders via Deep Latent-Variable Recurrent Neural Network

Granger causality analysis, as one of the most popular time series causality methods, has been widely used in the economics, neuroscience. However, unobserved confounders is a fundamental problem in the observational studies, which is still not solved for the non-linear Granger causality. The application works often deal with this problem in virtue of the proxy variables, who can be treated as a measure of the confounder with noise. But the proxy variables has been proved to be unreliable, because of the bias it may induce. In this paper, we try to "recover" the unobserved confounders for the Granger causality. We use a generative model with latent variable to build the relationship between the unobserved confounders and the observed variables(tested variable and the proxy variables). The posterior distribution of the latent variable is adopted to represent the confounders distribution, which can be sampled to get the estimated confounders. We adopt the variational autoencoder to estimate the intractable posterior distribution. The recurrent neural network is applied to build the temporal relationship in the data. We evaluate our method in the synthetic and semi-synthetic dataset. The result shows our estimated confounders has a better performance than the proxy variables in the non-linear Granger causality with multiple proxies in the semi-synthetic dataset. But the performances of the synthetic dataset and the different noise level of proxy seem terrible. Any advice can really help.

cs.LG↗

arXiv2019

Various methods for queue length and traffic volume estimation using probe vehicle trajectories

The rapid development of connected vehicle technology and the emergence of ride-hailing services have enabled the collection of a tremendous amount of probe vehicle trajectory data. Due to the large scale, the trajectory data have become a potential substitute for the widely used fixed-location sensors in terms of the performance measures of transportation networks. Specifically, for traffic volume and queue length estimation, most of the trajectory data based methods in the existing literature either require high market penetration of the probe vehicles to identify the shockwave or require the prior information about the queue length distribution and the penetration rate, which may not be feasible in the real world. To overcome the limitations of the existing methods, this paper proposes a series of novel methods based on probability theory. By exploiting the stopping positions of the probe vehicles in the queues, the proposed methods try to establish and solve a single-variable equation for the penetration rate of the probe vehicles. Once the penetration rate is obtained, it can be used to project the total queue length and the total traffic volume. The validation results using both simulation data and real-world data show that the methods would be accurate enough for assistance in performance measures and traffic signal control at intersections, even when the penetration rate of the probe vehicles is very low.

physics.soc-ph↗

arXiv2018

Hybrid topological evolution of multi-singularity vortex beams: Generalized nature for helical-Ince-Gaussian and Hermite-Laguerre-Gaussian modes

A generalized family of scalar structured Gaussian modes including helical-Ince--Gaussian (HIG) and Hermite--Laguerre--Gaussian (HLG) beams is presented with physical insight upon a hybrid topological evolution nature of multi-singularity vortex beams carrying orbital angular momentum (OAM). Considering the physical origins of intrinsic coordinates aberration and the Gouy phase shift, a closed-form expression is derived to characterize the general modes in astigmatic optical systems. Moreover, a graphical representation, Singularities Hybrid Evolution Nature (SHEN) sphere, is proposed to visualize the topological evolution of the multi-singularity beams, accommodating HLG, HIG and other typical subfamilies as characteristic curves on the sphere surface. The salient properties of SHEN sphere for describing the precise singularities splitting phenomena, exotic structured light fields, and Gouy phase shift are illustrated with adequate experimental verifications.

physics.optics↗

arXiv2018

Polygonal vortex beams in quasi-frequency-degenerate states

We originally demonstrate the vortex beams with patterns of closed polygons [namely polygonal vortex beams (PVBs)] generated by a quasi-frequency-degenerate (QFD) Yb:CALGO laser resonator with astigmatic transformation. The PVBs with peculiar patterns of triangular, square, and parallelogram shapes carrying large orbital angular momentums (OAMs) are theoretically investigated and experimentally obtained in the vicinity of the SU(2) degenerate states of laser resonator. The PVBs in QFD states are compared with the vortex beams with patterns of isolated spots arrays located on the triangle-, square-, and parallelogram-shaped routes [namely polygonalspots-array vortex beams (PSA-VBs)] under normal SU(2) degenerate states. Beam profile shape of PVB or PSA-VB and OAM can be controlled by adjusting the cavity length and the position of pump spot. The simulated and experimental results validate the performance of our method to generate PVB, which is of great potential for promoting novel technologies in particle trapping and beam shaping.

physics.optics↗

arXiv2018

Dual-wavelength vortex beam with high stability in diode-pumped Yb:CaGdAlO4 Laser

We present a stable dual-wavelength vortex beam carrying orbital angular momentum (OAM) with two spectral peaks separated by a few terahertz in diode-pumped Yb:CaGdAlO4 (CALGO) laser. The dual-wavelength spectrum is controlled by the pump power and off-axis loss in laser resonator, arising from the broad emission bandwidth of Yb:CALGO. The OAM beam is obtained by a pair of cylindrical lens that serves as an π/2 convertor for the high-order Hermite-Gaussian modes. The stability is verifed that the 1 h OAM beam with two spectral peaks at 1046.1 nm and 1057.2 nm (3.01 THz interval) can steadily operate for more than three hours. It has great potential for scaling the application for OAM beams in Terahertz spectroscopy, high-resolution interferometry, and so on.

physics.optics↗

arXiv2017

Why It Takes So Long to Connect to a WiFi Access Point

Today's WiFi networks deliver a large fraction of traffic. However, the performance and quality of WiFi networks are still far from satisfactory. Among many popular quality metrics (throughput, latency), the probability of successfully connecting to WiFi APs and the time cost of the WiFi connection set-up process are the two of the most critical metrics that affect WiFi users' experience. To understand the WiFi connection set-up process in real-world settings, we carry out measurement studies on $5$ million mobile users from $4$ representative cities associating with $7$ million APs in $0.4$ billion WiFi sessions, collected from a mobile "WiFi Manager" App that tops the Android/iOS App market. To the best of our knowledge, we are the first to do such large scale study on: how large the WiFi connection set-up time cost is, what factors affect the WiFi connection set-up process, and what can be done to reduce the WiFi connection set-up time cost. Based on the measurement analysis, we develop a machine learning based AP selection strategy that can significantly improve WiFi connection set-up performance, against the conventional strategy purely based on signal strength, by reducing the connection set-up failures from $33\%$ to $3.6\%$ and reducing $80\%$ time costs of the connection set-up processes by more than $10$ times.

cs.NI↗