arXiv subjects

Feng Zhu

Publications and source records attributed to Feng Zhu.

At least 127 records · Page 7Linked to original sources

arXiv2022

Protection of quantum evolutions under parity-time symmetric non-Hermitian Hamiltonians by dynamical decoupling

Parity-time (PT) symmetric non-Hermitian Hamiltonians bring about many novel features and interesting applications such as quantum gates faster than those in Hermitian systems, and topological state transfer. The performance of evolutions under $\mathcal{PT}$-symmetric Hamiltonians is degraded by the inevitable noise and errors due to system-environment interaction and experimental imperfections. In contrast to Hermitian Hamiltonians, the fluctuations in dissipative beams that are utilized to generate non-Hermitian contributions in the PT-symmetric Hamiltonians cause additional errors. Here we achieve the protection of PT-symmetric Hamiltonians against noise acting along the qubit's quantization axis by combining quantum evolutions with dynamical decoupling sequences. We demonstrate the performance of our method by numerical simulations. Realistic noise sources and parameters are chosen including: constant detuning error, time-varying detuning noise and dissipative-beam noise. The fidelities of the protected evolutions are well above the unprotected ones under all the above situations. Our work paves the way for further studies and applications of non-Hermitian $\mathcal{PT}$-symmetric physics in noisy quantum systems.

quant-ph↗

arXiv2022

Feature Erasing and Diffusion Network for Occluded Person Re-Identification

Occluded person re-identification (ReID) aims at matching occluded person images to holistic ones across different camera views. Target Pedestrians (TP) are usually disturbed by Non-Pedestrian Occlusions (NPO) and NonTarget Pedestrians (NTP). Previous methods mainly focus on increasing model's robustness against NPO while ignoring feature contamination from NTP. In this paper, we propose a novel Feature Erasing and Diffusion Network (FED) to simultaneously handle NPO and NTP. Specifically, NPO features are eliminated by our proposed Occlusion Erasing Module (OEM), aided by the NPO augmentation strategy which simulates NPO on holistic pedestrian images and generates precise occlusion masks. Subsequently, we Subsequently, we diffuse the pedestrian representations with other memorized features to synthesize NTP characteristics in the feature space which is achieved by a novel Feature Diffusion Module (FDM) through a learnable cross attention mechanism. With the guidance of the occlusion scores from OEM, the feature diffusion process is mainly conducted on visible body parts, which guarantees the quality of the synthesized NTP characteristics. By jointly optimizing OEM and FDM in our proposed FED network, we can greatly improve the model's perception ability towards TP and alleviate the influence of NPO and NTP. Furthermore, the proposed FDM only works as an auxiliary module for training and will be discarded in the inference phase, thus introducing little inference computational overhead. Experiments on occluded and holistic person ReID benchmarks demonstrate the superiority of FED over state-of-the-arts, where FED achieves 86.3% Rank-1 accuracy on Occluded-REID, surpassing others by at least 4.7%.

cs.CV↗

arXiv2022

Revisiting the Transferability of Supervised Pretraining: an MLP Perspective

The pretrain-finetune paradigm is a classical pipeline in visual learning. Recent progress on unsupervised pretraining methods shows superior transfer performance to their supervised counterparts. This paper revisits this phenomenon and sheds new light on understanding the transferability gap between unsupervised and supervised pretraining from a multilayer perceptron (MLP) perspective. While previous works focus on the effectiveness of MLP on unsupervised image classification where pretraining and evaluation are conducted on the same dataset, we reveal that the MLP projector is also the key factor to better transferability of unsupervised pretraining methods than supervised pretraining methods. Based on this observation, we attempt to close the transferability gap between supervised and unsupervised pretraining by adding an MLP projector before the classifier in supervised pretraining. Our analysis indicates that the MLP projector can help retain intra-class variation of visual features, decrease the feature distribution distance between pretraining and evaluation datasets, and reduce feature redundancy. Extensive experiments on public benchmarks demonstrate that the added MLP projector significantly boosts the transferability of supervised pretraining, e.g. +7.2% top-1 accuracy on the concept generalization task, +5.8% top-1 accuracy for linear evaluation on 12-domain classification tasks, and +0.8% AP on COCO object detection task, making supervised pretraining comparable or even better than unsupervised pretraining.

cs.CV↗

arXiv2022

Time Dependency, Data Flow, and Competitive Advantage

Data is fundamental to machine learning-based products and services and is considered strategic due to its externalities for businesses, governments, non-profits, and more generally for society. It is renowned that the value of organizations (businesses, government agencies and programs, and even industries) scales with the volume of available data. What is often less appreciated is that the data value in making useful organizational predictions will range widely and is prominently a function of data characteristics and underlying algorithms. In this research, our goal is to study how the value of data changes over time and how this change varies across contexts and business areas (e.g. next word prediction in the context of history, sports, politics). We focus on data from Reddit.com and compare the value's time-dependency across various Reddit topics (Subreddits). We make this comparison by measuring the rate at which user-generated text data loses its relevance to the algorithmic prediction of conversations. We show that different subreddits have different rates of relevance decline over time. Relating the text topics to various business areas of interest, we argue that competing in a business area in which data value decays rapidly alters strategies to acquire competitive advantage. When data value decays rapidly, access to a continuous flow of data will be more valuable than access to a fixed stock of data. In this kind of setting, improving user engagement and increasing user-base help creating and maintaining a competitive advantage.

cs.LG↗

arXiv2022

Quantum simulation of a general anti-PT-symmetric Hamiltonian with a trapped ion qubit

Non-Hermitian systems satisfying parity-time (PT) symmetry have aroused considerable interest owing to their exotic features. Anti-PT symmetry is an important counterpart of the PT symmetry, and has been studied in various classical systems. Although a Hamiltonian with anti-PT symmetry only differs from its PT-symmetric counterpart in a global $\pm$ i phase, the ways they change information and energy with the enivronment are completely different, suggesting an essential different dynamics in anti-PT-symmetric systems from their PT -symmetric variants. Moreover, theortical works have shown that qubits with anti-PT-symmetric Hamiltonians (anti-PT-symmetric qubits) have superior decoherence properties over Hermitian qubits, as well as PT-symmetric ones. So far, the observation of anti-PT symmetry in individual quantum systems remains elusive. Here, we implement an anti-PT -symmetric Hamiltonian of a single qubit in a single trapped ion by a designed microwave and optical control-pulse sequence. We characterize the anti-PT phase transition by mapping out the eigenvalues at different dissipation rates. The full inforamtion of the quantum state is also obtained by quantum state tomography. Our work allows quantum simulation of genuine open-system feature of an anti-PT-symmetric system, which paves the way for utilizing non-Hermitian properties for quantum information processing.

quant-ph↗

arXiv2022

A single-atom level mechano-optical transducer for ultrasensitive force sensing

Using light as a probe to detect a mechanical motion is one of the most successful experimental approaches in physics. The history of mechanical sensing based on the reflection, refraction and scattering of light dates back to the 16th century, where in the Cavendish experiment, the angle of rotation induced by the gravitational force is measured by the deflection of a light beam reflected from a mirror attached to the suspension. In modern science, mechano-optical transducers are such devices that could detect, measure and convert a force or displacement signal to an optical one, and are widely used for force detection. Especially, ultraweak force sensor with ultrahigh spatial resolution is highly demanded for detecting force anomaly in surface science, biomolecule imaging, and atomtronics. Here we show a novel scheme using a single trapped ion as a mechano-optical transduction. This method utilizes the force-induced micromotion, converting the micromotion to a time-resolved fluorescence signal, in which the ion's excess micromotion coupled to the Doppler shift of the scattered photons. We demonstrate the measurement sensitivity about 600 $\textrm{zN}/\sqrt{\textrm{Hz}}$ (1 $\textrm{zN} =10^{-21}$N). By alternating the detection laser beam in all three dimensions, the amplitude and the direction of a vector force can be precisely determined, constituting a 3D force sensor. This mechano-optical transducer provides high sensitivity with spatial resolution in single-atom level, enabling the applications in material industry and the search for possible exotic spin-dependent interactions that beyond the standard model.

physics.atom-ph↗

arXiv2022

Frequency stabilization of a 739 nm laser to an $I_2$ spectrum for trapped Ytterbium ions

We report on the frequency stabilization of a 739 nm Ti:sapphire laser to a hyperfine component of the $^{127}I_{2}$ B(1)-X(11) P(70) transition using acousto-optic modulation transfer spectroscopy (MTS). A frequency stability of $3.83\times 10^{-11}$ around 13 s averaging time is achieved when the laser frequency is stabilized. The observed hyperfine transition of the molecular iodine is an ideal frequency reference for locking the lasers used in experiments with trapped ytterbium ions, since its second harmonic frequency is the $^{2}S_{\frac{1}{2}}-^{2}P_{\frac{1}{2}}$ transition of the ytterbium ion at 369.5 nm. By investigating the line broadening effects due to the iodine vapor pressure and laser power, the locking is optimized to the theoretical signal to noise ratio (TSNR) of this iodine transition.

physics.atom-ph↗

arXiv2022

Relatively dominated representations from eigenvalue gaps and limit maps

Relatively dominated representations give a common generalization of geometrically finiteness in rank one on the one hand, and the Anosov condition which serves as a higher-rank analogue of convex cocompactness on the other. This note proves three results about these representations. Firstly, we remove the technical assumption of quadratic gaps involved in the original definition. Secondly, we give a characterization using eigenvalue gaps, providing a relative analogue of a result of Kassel-Potrie for Anosov representations. Thirdly, we formulate characterizations in terms of singular value or eigenvalue gaps combined with limit maps, in the spirit of Guéritaud-Guichard-Kassel-Wienhard for Anosov representations, and use them to show that inclusion representations of certain groups playing weak ping-pong and positive representations in the sense of Fock-Goncharov are relatively dominated.

math.GR↗

arXiv2022

MixMix: All You Need for Data-Free Compression Are Feature and Data Mixing

User data confidentiality protection is becoming a rising challenge in the present deep learning research. Without access to data, conventional data-driven model compression faces a higher risk of performance degradation. Recently, some works propose to generate images from a specific pretrained model to serve as training data. However, the inversion process only utilizes biased feature statistics stored in one model and is from low-dimension to high-dimension. As a consequence, it inevitably encounters the difficulties of generalizability and inexact inversion, which leads to unsatisfactory performance. To address these problems, we propose MixMix based on two simple yet effective techniques: (1) Feature Mixing: utilizes various models to construct a universal feature space for generalized inversion; (2) Data Mixing: mixes the synthesized images and labels to generate exact label information. We prove the effectiveness of MixMix from both theoretical and empirical perspectives. Extensive experiments show that MixMix outperforms existing methods on the mainstream compression tasks, including quantization, knowledge distillation, and pruning. Specifically, MixMix achieves up to 4% and 20% accuracy uplift on quantization and pruning, respectively, compared to existing data-free compression work.

cs.LG↗

arXiv2022

GPS: A Policy-driven Sampling Approach for Graph Representation Learning

Graph representation learning has drawn increasing attention in recent years, especially for learning the low dimensional embedding at both node and graph level for classification and recommendations tasks. To enable learning the representation on the large-scale graph data in the real world, numerous research has focused on developing different sampling strategies to facilitate the training process. Herein, we propose an adaptive Graph Policy-driven Sampling model (GPS), where the influence of each node in the local neighborhood is realized through the adaptive correlation calculation. Specifically, the selections of the neighbors are guided by an adaptive policy algorithm, contributing directly to the message aggregation, node embedding updating, and graph level readout steps. We then conduct comprehensive experiments against baseline methods on graph classification tasks from various perspectives. Our proposed model outperforms the existing ones by 3%-8% on several vital benchmarks, achieving state-of-the-art performance in real-world datasets.

cs.LG↗

arXiv2022

Adaptive Worker Grouping For Communication-Efficient and Straggler-Tolerant Distributed SGD

Wall-clock convergence time and communication load are key performance metrics for the distributed implementation of stochastic gradient descent (SGD) in parameter server settings. Communication-adaptive distributed Adam (CADA) has been recently proposed as a way to reduce communication load via the adaptive selection of workers. CADA is subject to performance degradation in terms of wall-clock convergence time in the presence of stragglers. This paper proposes a novel scheme named grouping-based CADA (G-CADA) that retains the advantages of CADA in reducing the communication load, while increasing the robustness to stragglers at the cost of additional storage at the workers. G-CADA partitions the workers into groups of workers that are assigned the same data shards. Groups are scheduled adaptively at each iteration, and the server only waits for the fastest worker in each selected group. We provide analysis and experimental results to elaborate the significant gains on the wall-clock time, as well as communication load and computation load, of G-CADA over other benchmark schemes.

cs.IT↗

arXiv2021

Offline Planning and Online Learning under Recovering Rewards

Motivated by emerging applications such as live-streaming e-commerce, promotions and recommendations, we introduce and solve a general class of non-stationary multi-armed bandit problems that have the following two features: (i) the decision maker can pull and collect rewards from up to $K\,(\ge 1)$ out of $N$ different arms in each time period; (ii) the expected reward of an arm immediately drops after it is pulled, and then non-parametrically recovers as the arm's idle time increases. With the objective of maximizing the expected cumulative reward over $T$ time periods, we design a class of ``Purely Periodic Policies'' that jointly set a period to pull each arm. For the proposed policies, we prove performance guarantees for both the offline problem and the online problems. For the offline problem when all model parameters are known, the proposed periodic policy obtains an approximation ratio that is at the order of $1-\mathcal O(1/\sqrt{K})$, which is asymptotically optimal when $K$ grows to infinity. For the online problem when the model parameters are unknown and need to be dynamically learned, we integrate the offline periodic policy with the upper confidence bound procedure to construct on online policy. The proposed online policy is proved to approximately have $\widetilde{\mathcal O}(N\sqrt{T})$ regret against the offline benchmark. Our framework and policy design may shed light on broader offline planning and online learning applications with non-stationary and recovering rewards.

stat.ML↗

arXiv2021

DronePose: The identification, segmentation, and orientation detection of drones via neural networks

The growing ubiquity of drones has raised concerns over the ability of traditional air-space monitoring technologies to accurately characterise such vehicles. Here, we present a CNN using a decision tree and ensemble structure to fully characterise drones in flight. Our system determines the drone type, orientation (in terms of pitch, roll, and yaw), and performs segmentation to classify different body parts (engines, body, and camera). We also provide a computer model for the rapid generation of large quantities of accurately labelled photo-realistic training data and demonstrate that this data is of sufficient fidelity to allow the system to accurately characterise real drones in flight. Our network will provide a valuable tool in the image processing chain where it may build upon existing drone detection technologies to provide complete drone characterisation over wide areas.

cs.CV↗

arXiv2021

Self-distillation with Batch Knowledge Ensembling Improves ImageNet Classification

The recent studies of knowledge distillation have discovered that ensembling the "dark knowledge" from multiple teachers or students contributes to creating better soft targets for training, but at the cost of significantly more computations and/or parameters. In this work, we present BAtch Knowledge Ensembling (BAKE) to produce refined soft targets for anchor images by propagating and ensembling the knowledge of the other samples in the same mini-batch. Specifically, for each sample of interest, the propagation of knowledge is weighted in accordance with the inter-sample affinities, which are estimated on-the-fly with the current network. The propagated knowledge can then be ensembled to form a better soft target for distillation. In this way, our BAKE framework achieves online knowledge ensembling across multiple samples with only a single network. It requires minimal computational and memory overhead compared to existing knowledge ensembling methods. Extensive experiments demonstrate that the lightweight yet effective BAKE consistently boosts the classification performance of various architectures on multiple datasets, e.g., a significant +0.7% gain of Swin-T on ImageNet with only +1.5% computational overhead and zero additional parameters. BAKE does not only improve the vanilla baselines, but also surpasses the single-network state-of-the-arts on all the benchmarks.

cs.CV↗

arXiv2021

Ergodicity and equidistribution in Hilbert geometry

We show that dynamical and counting results characteristic of negatively-curved Riemannian geometry, or more generally CAT($-1$) or rank-one CAT($0$) spaces, also hold for rank-one properly convex projective structures, equipped with their Hilbert metrics, admitting finite Sullivan measures built from appropriate conformal densities. In particular, this includes geometrically finite convex projective structures. More specifically, with respect to the Sullivan measure, the Hilbert geodesic flow is strongly mixing, and orbits and primitive closed geodesics equidistribute, allowing us to asymptotically enumerate these objects.

math.DS↗

arXiv2021

Deep inference of latent dynamics with spatio-temporal super-resolution using selective backpropagation through time

Modern neural interfaces allow access to the activity of up to a million neurons within brain circuits. However, bandwidth limits often create a trade-off between greater spatial sampling (more channels or pixels) and the temporal frequency of sampling. Here we demonstrate that it is possible to obtain spatio-temporal super-resolution in neuronal time series by exploiting relationships among neurons, embedded in latent low-dimensional population dynamics. Our novel neural network training strategy, selective backpropagation through time (SBTT), enables learning of deep generative models of latent dynamics from data in which the set of observed variables changes at each time step. The resulting models are able to infer activity for missing samples by combining observations with learned latent dynamics. We test SBTT applied to sequential autoencoders and demonstrate more efficient and higher-fidelity characterization of neural population dynamics in electrophysiological and calcium imaging data. In electrophysiology, SBTT enables accurate inference of neuronal population dynamics with lower interface bandwidths, providing an avenue to significant power savings for implanted neuroelectronic interfaces. In applications to two-photon calcium imaging, SBTT accurately uncovers high-frequency temporal structure underlying neural population activity, substantially outperforming the current state-of-the-art. Finally, we demonstrate that performance could be further improved by using limited, high-bandwidth sampling to pretrain dynamics models, and then using SBTT to adapt these models for sparsely-sampled data.

cs.LG↗

arXiv2021

Statistical CSI-Based Transmission Design for Reconfigurable Intelligent Surface-aided Massive MIMO Systems with Hardware Impairments

We consider a reconfigurable intelligent surface (RIS)-aided massive multi-user multiple-input multiple-output (MIMO) communication system with transceiver hardware impairments (HWIs) and RIS phase noise. Different from the existing contributions, the phase shifts of the RIS are designed based on the long-term angle informations. Firstly, an approximate analytical expression of the uplink achievable rate is derived. Then, we use genetic algorithm (GA) to maximize the sum rate and the minimum date rate. Finally, we show that it is crucial to take HWIs into account when designing the phase shift of RIS.

cs.IT↗

arXiv2021

Crashworthiness design of 3D lattice-structure filled thin-walled tubes based on data mining

Lattice structures and thin-walled tubes are two types of energy-absorbers widely studied and applied in engineering practice. In this study, a new type of lattice-structure filled thin-walled tube (LFT) was proposed. In this new type of LFT, a BCC-Z lattice structure was filled into a square thin-walled tube. Then using data mining, a 3-D geometric design with five design variables was conducted on this new LFT. Using Latin Hypercubic sampling algorithm, 150 design cases were generated. Numerical models were then developed to simulate their crush behavior, and the simulation dataset was used for data mining. The results showed that (1) Filling the BBC-Z lattice structure into a thin-walled tube can significantly improve the energy absorption (EA) capacity of the structure. (2) The decision trees generated in the data mining process indicated that the rod diameter d of lattice structure is the key design variable that has most significant impact on EA, followed by m and n. (3) The design rules to build LFTs with high EA efficiency (SEA>=16 kJ/kg and CFE>=45%), high total EA (SEA>=16 kJ/kg and EA>=6 kJ), and lightweight (SEA>=16 kJ/kg and Mass<=0.45 kg) were obtained from decision trees. The ideal configurations of LFT corresponding to these three objectives are: d>2 mm, n>2 and m>3 for high EA efficiency; d>2 mm, n>2 and m>3 for high total EA; and d>2 mm, n>2, m<=4 and t<=1.7 mm for lightweight.

cs.DC↗