arXiv subjects

Donghao Li

Publications and source records attributed to Donghao Li.

34 records · Page 2Linked to original sources

arXiv2023

NeuroMixGDP: A Neural Collapse-Inspired Random Mixup for Private Data Release

Privacy-preserving data release algorithms have gained increasing attention for their ability to protect user privacy while enabling downstream machine learning tasks. However, the utility of current popular algorithms is not always satisfactory. Mixup of raw data provides a new way of data augmentation, which can help improve utility. However, its performance drastically deteriorates when differential privacy (DP) noise is added. To address this issue, this paper draws inspiration from the recently observed Neural Collapse (NC) phenomenon, which states that the last layer features of a neural network concentrate on the vertices of a simplex as Equiangular Tight Frame (ETF). We propose a scheme to mixup the Neural Collapse features to exploit the ETF simplex structure and release noisy mixed features to enhance the utility of the released data. By using Gaussian Differential Privacy (GDP), we obtain an asymptotic rate for the optimal mixup degree. To further enhance the utility and address the label collapse issue when the mixup degree is large, we propose a Hierarchical sampling method to stratify the mixup samples on a small number of classes. This method remarkably improves utility when the number of classes is large. Extensive experiments demonstrate the effectiveness of our proposed method in protecting against attacks and improving utility. In particular, our approach shows significantly improved utility compared to directly training classification networks with DPSGD on CIFAR100 and MiniImagenet datasets, highlighting the benefits of using privacy-preserving data release. We release reproducible code in https://github.com/Lidonghao1996/NeuroMixGDP.

cs.LG↗

arXiv2023

ChatCounselor: A Large Language Models for Mental Health Support

This paper presents ChatCounselor, a large language model (LLM) solution designed to provide mental health support. Unlike generic chatbots, ChatCounselor is distinguished by its foundation in real conversations between consulting clients and professional psychologists, enabling it to possess specialized knowledge and counseling skills in the field of psychology. The training dataset, Psych8k, was constructed from 260 in-depth interviews, each spanning an hour. To assess the quality of counseling responses, the counseling Bench was devised. Leveraging GPT-4 and meticulously crafted prompts based on seven metrics of psychological counseling assessment, the model underwent evaluation using a set of real-world counseling questions. Impressively, ChatCounselor surpasses existing open-source models in the counseling Bench and approaches the performance level of ChatGPT, showcasing the remarkable enhancement in model capability attained through high-quality domain-specific data.

cs.CL↗

arXiv2023

Robot motor learning shows emergence of frequency-modulated, robust swimming with an invariant Strouhal-number

Fish locomotion emerges from a diversity of interactions among deformable structures, surrounding fluids and neuromuscular activations, i.e., fluid-structure interactions (FSI) controlled by fish's motor systems. Previous studies suggested that such motor-controlled FSI may possess embodied traits. However, their implications in motor learning, neuromuscular control, gait generation, and swimming performance remain to be uncovered. Using robot models, we studied how swimming behaviours emerged from the FSI and the embodied traits. We developed modular robots with various designs and used Central Pattern Generators (CPGs) to control the torque acting on robot body. We used reinforcement learning to learn CPG parameters to maximize the swimming speed. The results showed that motor frequency converged faster than other parameters, and the emergent swimming gaits were robust against disruptions applied to motor control. For all robots and frequencies tested, swimming speed was proportional to the mean undulation velocity of body and caudal-fin combined, yielding an invariant, undulation-based Strouhal number. The Strouhal number also revealed two fundamental classes of undulatory swimming in both biological and robotic fishes. The robot actuators also demonstrated diverse functions as motors, virtual springs, and virtual masses. These results provide novel insights into the embodied traits of motor-controlled FSI for fish-inspired locomotion.

cs.RO↗

arXiv2023

Near-optimal Conservative Exploration in Reinforcement Learning under Episode-wise Constraints

This paper investigates conservative exploration in reinforcement learning where the performance of the learning agent is guaranteed to be above a certain threshold throughout the learning process. It focuses on the tabular episodic Markov Decision Process (MDP) setting that has finite states and actions. With the knowledge of an existing safe baseline policy, an algorithm termed as StepMix is proposed to balance the exploitation and exploration while ensuring that the conservative constraint is never violated in each episode with high probability. StepMix features a unique design of a mixture policy that adaptively and smoothly interpolates between the baseline policy and the optimistic policy. Theoretical analysis shows that StepMix achieves near-optimal regret order as in the constraint-free setting, indicating that obeying the stringent episode-wise conservative constraint does not compromise the learning performance. Besides, a randomization-based EpsMix algorithm is also proposed and shown to achieve the same performance as StepMix. The algorithm design and theoretical analysis are further extended to the setting where the baseline policy is not given a priori but must be learned from an offline dataset, and it is proved that similar conservative guarantee and regret can be achieved if the offline dataset is sufficiently large. Experiment results corroborate the theoretical analysis and demonstrate the effectiveness of the proposed conservative exploration strategies.

cs.LG↗

arXiv2023

Random Smoothing Regularization in Kernel Gradient Descent Learning

Random smoothing data augmentation is a unique form of regularization that can prevent overfitting by introducing noise to the input data, encouraging the model to learn more generalized features. Despite its success in various applications, there has been a lack of systematic study on the regularization ability of random smoothing. In this paper, we aim to bridge this gap by presenting a framework for random smoothing regularization that can adaptively and effectively learn a wide range of ground truth functions belonging to the classical Sobolev spaces. Specifically, we investigate two underlying function spaces: the Sobolev space of low intrinsic dimension, which includes the Sobolev space in $D$-dimensional Euclidean space or low-dimensional sub-manifolds as special cases, and the mixed smooth Sobolev space with a tensor structure. By using random smoothing regularization as novel convolution-based smoothing kernels, we can attain optimal convergence rates in these cases using a kernel gradient descent algorithm, either with early stopping or weight decay. It is noteworthy that our estimator can adapt to the structural assumptions of the underlying data and avoid the curse of dimensionality. This is achieved through various choices of injected noise distributions such as Gaussian, Laplace, or general polynomial noises, allowing for broad adaptation to the aforementioned structural assumptions of the underlying data. The convergence rate depends only on the effective dimension, which may be significantly smaller than the actual data dimension. We conduct numerical experiments on simulated data to validate our theoretical results.

stat.ML↗

arXiv2022

Exploring Structural Sparsity of Deep Networks via Inverse Scale Spaces

The great success of deep neural networks is built upon their over-parameterization, which smooths the optimization landscape without degrading the generalization ability. Despite the benefits of over-parameterization, a huge amount of parameters makes deep networks cumbersome in daily life applications. Though techniques such as pruning and distillation are developed, they are expensive in fully training a dense network as backward selection methods, and there is still a void on systematically exploring forward selection methods for learning structural sparsity in deep networks. To fill in this gap, this paper proposes a new approach based on differential inclusions of inverse scale spaces, which generate a family of models from simple to complex ones along the dynamics via coupling a pair of parameters, such that over-parameterized deep models and their structural sparsity can be explored simultaneously. This kind of differential inclusion scheme has a simple discretization, dubbed Deep structure splitting Linearized Bregman Iteration (DessiLBI), whose global convergence in learning deep networks could be established under the Kurdyka-Lojasiewicz framework. Experimental evidence shows that our method achieves comparable and even better performance than the competitive optimizers in exploring the sparse structure of several widely used backbones on the benchmark datasets. Remarkably, with early stopping, our method unveils `winning tickets' in early epochs: the effective sparse network structures with comparable test accuracy to fully trained over-parameterized models, that are further transferable to similar alternative tasks. Furthermore, our method is able to grow networks efficiently with adaptive filter configurations, demonstrating a good performance with much less computational cost. Codes and models can be downloaded at {https://github.com/DessiLBI2020/DessiLBI}.

cs.LG↗

arXiv2021

Effects of Design and Hydrodynamic Parameters on Optimized Swimming for Simulated, Fish-inspired Robots

In this work we developed a mathematical model and a simulation platform for a fish-inspired robotic template, namely Magnetic, Modular, Undulatory Robotics ($μ$Bots). Through this platform, we systematically explored the effects of design and fluid parameters on the swimming performance via reinforcement learning. The mathematical model was composed of two interacting subsystems, the robot dynamics and the hydrodynamics, and the hydrodynamic model consisted of reactive components (added-mass and pressure forces) and resistive components (drag and friction forces), which were then nondimensionalized for deriving key "control parameters" of robot-fluid interaction. The $μ$Bot was actuated via magnetic actuators controlled with harmonic voltage signals, which were optimized via EM-based Policy Hyper Parameter Exploration (EPHE) to maximize swimming speed. By varying the control parameters, total 36 cases with different robot template variations (Number of Actuation (NoA) and stiffness) and hydrodynamic parameters were simulated and optimized via EPHE. Results showed that wavelength of optimized gaits (i.e., traveling wave along body) was independent of template variations and hydrodynamic parameters. Higher NoA yielded higher speed but lower speed per body length however with diminishing gain and lower speed per body length. Body and caudal-fin gait dynamics were dominated by the interaction among fluid added-mass, spring, and actuation torque, with negligible contribution from fluid resistive drag. In contrast, thrust generation was dominated by pressure force acting on caudal fin, as steady swimming resulted from a balance between resistive force and pressure force, with minor contributions from added-mass and body drag forces. Therefore, added-mass force only indirectly affected the thrust generation and swimming speed via the caudal fin dynamics.

cs.RO↗

arXiv2021

Particle filter re-detection for visual tracking via correlation filters

Most of the correlation filter based tracking algorithms can achieve good performance and maintain fast computational speed. However, in some complicated tracking scenes, there is a fatal defect that causes the object to be located inaccurately. In order to address this problem, we propose a particle filter redetection based tracking approach for accurate object localization. During the tracking process, the kernelized correlation filter (KCF) based tracker locates the object by relying on the maximum response value of the response map; when the response map becomes ambiguous, the KCF tracking result becomes unreliable. Our method can provide more candidates by particle resampling to detect the object accordingly. Additionally, we give a new object scale evaluation mechanism, which merely considers the differences between the maximum response values in consecutive frames. Extensive experiments on OTB2013 and OTB2015 datasets demonstrate that the proposed tracker performs favorably in relation to the state-of-the-art methods.

cs.CV↗

arXiv2020

The effects of nonlinear damping on degenerate parametric amplification

This paper considers the dynamic response of a single degree of freedom system with nonlinear stiffness and nonlinear damping that is subjected to both resonant direct excitation and resonant parametric excitation, with a general phase between the two. This generalizes and expands on previous studies of nonlinear effects on parametric amplification, notably by including the effects of nonlinear damping, which is commonly observed in a large variety of systems, including micro- and nano-scale resonators. Using the method of averaging, a thorough parameter study is carried out that describes the effects of the amplitudes and relative phase of the two forms of excitation. The effects of nonlinear damping on the parametric gain are first derived. The transitions among various topological forms of the frequency response curves, which can include isolae, dual peaks, and loops, are determined, and bifurcation analyses in parameter spaces of interest are carried out. In general, these results provide a complete picture of the system response and allow one to select drive conditions of interest that avoid bistability while providing maximum amplitude gain, maximum phase sensitivity, or a flat resonant peak, in systems with nonlinear damping.

cond-mat.mes-hall↗

arXiv2020

DessiLBI: Exploring Structural Sparsity of Deep Networks via Differential Inclusion Paths

Over-parameterization is ubiquitous nowadays in training neural networks to benefit both optimization in seeking global optima and generalization in reducing prediction error. However, compressive networks are desired in many real world applications and direct training of small networks may be trapped in local optima. In this paper, instead of pruning or distilling over-parameterized models to compressive ones, we propose a new approach based on differential inclusions of inverse scale spaces. Specifically, it generates a family of models from simple to complex ones that couples a pair of parameters to simultaneously train over-parameterized deep models and structural sparsity on weights of fully connected and convolutional layers. Such a differential inclusion scheme has a simple discretization, proposed as Deep structurally splitting Linearized Bregman Iteration (DessiLBI), whose global convergence analysis in deep learning is established that from any initializations, algorithmic iterations converge to a critical point of empirical risks. Experimental evidence shows that DessiLBI achieve comparable and even better performance than the competitive optimizers in exploring the structural sparsity of several widely used backbones on the benchmark datasets. Remarkably, with early stopping, DessiLBI unveils "winning tickets" in early epochs: the effective sparse structure with comparable test accuracy to fully trained over-parameterized models.

cs.CV↗

arXiv2020

Experimental realization of spin-tensor momentum coupling in ultracold Fermi gases

We experimentally realize the spin-tensor momentum coupling (STMC) using the three ground Zeeman states coupled by three Raman laser beams in ultracold atomic system of $^{40}$K Fermi atoms. This new type of STMC consists of two bright-state bands as a regular spin-orbit coupled spin-1/2 system and one dark-state middle band. Using radio-frequency spin-injection spectroscopy, we investigate the energy band of STMC. It is demonstrated that the middle state is a dark state in the STMC system. The realized energy band of STMC may open the door for further exploring exotic quantum matters.

cond-mat.quant-gas↗

arXiv2019

$S^{2}$-LBI: Stochastic Split Linearized Bregman Iterations for Parsimonious Deep Learning

This paper proposes a novel Stochastic Split Linearized Bregman Iteration ($S^{2}$-LBI) algorithm to efficiently train the deep network. The $S^{2}$-LBI introduces an iterative regularization path with structural sparsity. Our $S^{2}$-LBI combines the computational efficiency of the LBI, and model selection consistency in learning the structural sparsity. The computed solution path intrinsically enables us to enlarge or simplify a network, which theoretically, is benefited from the dynamics property of our $S^{2}$-LBI algorithm. The experimental results validate our $S^{2}$-LBI on MNIST and CIFAR-10 dataset. For example, in MNIST, we can either boost a network with only 1.5K parameters (1 convolutional layer of 5 filters, and 1 FC layer), achieves 98.40\% recognition accuracy; or we simplify $82.5\%$ of parameters in LeNet-5 network, and still achieves the 98.47\% recognition accuracy. In addition, we also have the learning results on ImageNet, which will be added in the next version of our report.

stat.ML↗

arXiv2017

Universal Feature in Optical Control of a p-wave Feshbach Resonance

In this Letter we report the experimental results on optical control of a p-wave Feshbach resonance, by utilizing a laser driven bound-to-bound transition to shift the energy of closed channel molecule. The magnetic field location for p-wave resonance as a function of laser detuning can be captured by a simple formula with essentially one parameter, which describes how sensitive the resonance depends on the laser detuning. The key result of this work is to demonstrate, both experimentally and theoretically, that the ratio between this parameter for $m=0$ resonance and that for $m=\pm 1$ resonance, to large extent, is universal. We also show that this optical control can create intriguing situations where interesting few- and many-body physics can occurs, such as a p-wave resonance overlapping with an s-wave resonance or three p-wave resonances being degenerate.

cond-mat.quant-gas↗

arXiv2017

Observation of Floquet band topology change in driven ultracold Fermi gases

Periodic driving of a quantum system can significantly alter its energy bands and even change the band topology, opening a completely new avenue for engineering novel quantum matter. Although important progress has been made recently in measuring topological properties of Floquet bands in different systems, direct experimental measurement of Floquet band dispersions and their topology change is still demanding. Here we directly measure Floquet band dispersions in a periodically driven spin-orbit coupled ultracold Fermi gas. Using spin injection radio-frequency spectroscopy, we observe that the Dirac point originating from two dimensional spin-orbit coupling can be manipulated to emerge at the lowest or highest two dressed bands by fast modulating Raman laser frequencies, demonstrating topological change of Floquet bands. Our work will provide a powerful tool for understanding fundamental Floquet physics as well as engineering exotic topological quantum matter.

cond-mat.quant-gas↗

arXiv2016

Experimental observation of a topological band gap opening in ultracold Fermi gases with two-dimensional spin-orbit coupling

The recent experimental realization of synthetic spin-orbit coupling (SOC) opens a new avenue for exploring novel quantum states with ultracold atoms. However, in experiments for generating two-dimensional SOC (e.g., Rashba type), a perpendicular Zeeman field, which opens a band gap at the Dirac point and induces many topological phenomena, is still lacking. Here we theoretically propose and experimentally realize a simple scheme for generating two-dimension SOC and a perpendicular Zeeman field simultaneously in ultracold Fermi gases by tuning the polarization of three Raman lasers that couple three hyperfine ground states of atoms. The resulting band gap opening at the Dirac point is probed using spin injection radio-frequency spectroscopy. Our observation may pave the way for exploring topological transport and topological superfluids with exotic Majorana and Weyl fermion excitations in ultracold atoms.

cond-mat.quant-gas↗

arXiv2015

Experimental realization of a two-dimensional synthetic spin-orbit coupling in ultracold Fermi gases

Spin-orbit coupling (SOC) is central to many physical phenomena, including fine structures of atomic spectra and quantum topological matters. Whereas SOC is in general fixed in a physical system, atom-laser interaction provides physicists a unique means to create and control synthetic SOC for ultracold atoms \cite{Dalibard}. Though significant experimental progresses have been made, a bottleneck in current studies is the lack of a two-dimensional (2D) synthetic SOC, which is crucial for realizing high-dimensional topological matters. Here, we report the experimental realization of 2D SOC in ultracold $^{40}$K Fermi gases using three lasers, each of which dresses one atomic hyperfine spin state. Through spin injection radio-frequency (rf) spectroscopy, we probe the spin-resolved energy dispersions of dressed atoms, and observe a highly controllable Dirac point created by the 2D SOC. Our work paves the way for exploring high-dimensional topological matters in ultracold atoms using Raman schemes.

cond-mat.quant-gas↗