arXiv subjects

Minwoo Lee

Publications and source records attributed to Minwoo Lee.

At least 37 records · Page 2Linked to original sources

arXiv2023

Influence of Organic Spacer Cation on Dark Excitons in 2D Perovskites

The organic spacer cation plays a crucial role in determining the exciton fine structure in two-dimensional (2D) perovskites. Here, we use low-temperature magneto-optical spectroscopy to gain insight into the influence of the organic spacer on dark excitons in Ruddlesden-Popper (RP) perovskites. We show that by using modest magnetic field strengths (<1.5 T), the spin-forbidden dark exciton state can be identified and its emission properties significantly modulated through the application of in-plane magnetic fields, up to temperatures of 15 K. At low temperatures, an increase in collected photoluminescence efficiency of >30% is demonstrated, signifying the critical role of the dark exciton state for light-emitting applications of 2D perovskites. The exciton fine structure and the degree of magnetic-field-induced mixing are impacted by the choice of organic spacer cation, with 4-methoxyphenethylammonium (MeO-PEA) showing the largest effect due to larger bright-dark exciton splitting. Our results suggest that dark excitons preferentially form biexcitons depending on the choice of spacer. We distinguish between interior (bulk) and surface dark-exciton emission, showing that bright-dark exciton splitting differs between the interior and surface. Our results emphasize the significance of the organic spacer cation in controlling the exciton fine structure in 2D perovskites and have important implications for the development of optoelectronic technology based on 2D perovskites.

cond-mat.mtrl-sci↗

arXiv2023

A numerically efficient output-only system-identification framework for stochastically forced self-sustained oscillators

Self-sustained oscillations are ubiquitous in nature and engineering. In this paper, we propose a novel output-only system-identification framework for identifying the system parameters of a self-sustained oscillator affected by Gaussian white noise. A Langevin model that characterizes the self-sustained oscillator is postulated, and the corresponding Fokker--Planck equation is derived from stochastic averaging. From the drift and diffusion terms of the Fokker--Planck equation, unknown parameters of the system are identified. We develop a numerically efficient algorithm for enhancing the accuracy of parameter identification. In particular, a modified Levenberg--Marquardt optimization algorithm tailored to output-only system identification is introduced. The proposed framework is demonstrated on both numerical and experimental oscillators with varying system parameters that develop into self-sustained oscillations. The results show that the computational cost required for performing the system identification is dramatically reduced by using the proposed framework. Also, system parameters that were difficult to be extracted with the existing method could be efficiently computed with the system identification method developed in this study. Pertaining to the robustness and computational efficiency of the presented framework, this study can contribute to an accurate and fast diagnosis of dynamical systems under stochastic forcing.

eess.SY↗

arXiv2023

GaitSADA: Self-Aligned Domain Adaptation for mmWave Gait Recognition

mmWave radar-based gait recognition is a novel user identification method that captures human gait biometrics from mmWave radar return signals. This technology offers privacy protection and is resilient to weather and lighting conditions. However, its generalization performance is yet unknown and limits its practical deployment. To address this problem, in this paper, a non-synthetic dataset is collected and analyzed to reveal the presence of spatial and temporal domain shifts in mmWave gait biometric data, which significantly impacts identification accuracy. To mitigate this issue, a novel self-aligned domain adaptation method called GaitSADA is proposed. GaitSADA improves system generalization performance by using a two-stage semi-supervised model training approach. The first stage employs semi-supervised contrastive learning to learn a compact gait representation from both source and target domain data, aligning source-target domain distributions implicitly. The second stage uses semi-supervised consistency training with centroid alignment to further close source-target domain gap by pseudo-labelling the target-domain samples, clustering together the samples belonging to the same class but from different domains, and pushing the class centroid close to the weight vector of each class. Experiments show that GaitSADA outperforms representative domain adaptation methods with an improvement ranging from 15.41\% to 26.32\% on average accuracy in low data regimes. Code and dataset will be available at https://exitudio.github.io/GaitSADA

cs.CV↗

arXiv2023

Enabling Kernel Bypass Networking on gem5

Full-system simulation of computer systems is critical to capture the complex interplay between various hardware and software components in future systems. Modeling the network subsystem is indispensable to the fidelity of the full-system simulation due to the increasing importance of scale-out systems. The network software stack has undergone major changes over the last decade, and kernel-bypass networking stacks and data-plane networks are rapidly replacing the conventional kernel network stack. Nevertheless, the current state-of-the-art architectural simulators still use kernel networking which precludes realistic network application scenarios. In this work, we enable kernel bypass networking stack on gem5, the state-of-the-art full-system architectural simulator. We extend gem5's NIC hardware model and device driver to enable the support for userspace device drivers to run the DPDK framework. We also implement a network load generator hardware model in gem5 to generate various traffic patterns and perform per-packet timestamp and latency measurements without introducing packet loss. Our experimental results show that DPDK's simulated network bandwidth scales with the number of cores and NIC ports. As two use cases, we analyze the sensitivity of (1) network performance to several microarchitectural parameters, and (2) direct cache access (DCA) technology to DPDK burst size.

cs.AR↗

arXiv2023

Error-related Potential Variability: Exploring the Effects on Classification and Transferability

Brain-Computer Interfaces (BCI) have allowed for direct communication from the brain to external applications for the automatic detection of cognitive processes such as error recognition. Error-related potentials (ErrPs) are a particular brain signal elicited when one commits or observes an erroneous event. However, due to the noisy properties of the brain and recording devices, ErrPs vary from instance to instance as they are combined with an assortment of other brain signals, biological noise, and external noise, making the classification of ErrPs a non-trivial problem. Recent works have revealed particular cognitive processes such as awareness, embodiment, and predictability that contribute to ErrP variations. In this paper, we explore the performance of classifier transferability when trained on different ErrP variation datasets generated by varying the levels of awareness and embodiment for a given task. In particular, we look at transference between observational and interactive ErrP categories when elicited by similar and differing tasks. Our empirical results provide an exploratory analysis into the ErrP transferability problem from a data perspective.

cs.HC↗

arXiv2022

GaitMixer: Skeleton-based Gait Representation Learning via Wide-spectrum Multi-axial Mixer

Most existing gait recognition methods are appearance-based, which rely on the silhouettes extracted from the video data of human walking activities. The less-investigated skeleton-based gait recognition methods directly learn the gait dynamics from 2D/3D human skeleton sequences, which are theoretically more robust solutions in the presence of appearance changes caused by clothes, hairstyles, and carrying objects. However, the performance of skeleton-based solutions is still largely behind the appearance-based ones. This paper aims to close such performance gap by proposing a novel network model, GaitMixer, to learn more discriminative gait representation from skeleton sequence data. In particular, GaitMixer follows a heterogeneous multi-axial mixer architecture, which exploits the spatial self-attention mixer followed by the temporal large-kernel convolution mixer to learn rich multi-frequency signals in the gait feature maps. Experiments on the widely used gait database, CASIA-B, demonstrate that GaitMixer outperforms the previous SOTA skeleton-based methods by a large margin while achieving a competitive performance compared with the representative appearance-based solutions. Code will be available at https://github.com/exitudio/gaitmixer

cs.CV↗

arXiv2022

Privacy Enhancement for Cloud-Based Few-Shot Learning

Requiring less data for accurate models, few-shot learning has shown robustness and generality in many application domains. However, deploying few-shot models in untrusted environments may inflict privacy concerns, e.g., attacks or adversaries that may breach the privacy of user-supplied data. This paper studies the privacy enhancement for the few-shot learning in an untrusted environment, e.g., the cloud, by establishing a novel privacy-preserved embedding space that preserves the privacy of data and maintains the accuracy of the model. We examine the impact of various image privacy methods such as blurring, pixelization, Gaussian noise, and differentially private pixelization (DP-Pix) on few-shot image classification and propose a method that learns privacy-preserved representation through the joint loss. The empirical results show how privacy-performance trade-off can be negotiated for privacy-enhanced few-shot learning.

cs.LG↗

arXiv2022

Learning from Few Examples: A Summary of Approaches to Few-Shot Learning

Few-Shot Learning refers to the problem of learning the underlying pattern in the data just from a few training samples. Requiring a large number of data samples, many deep learning solutions suffer from data hunger and extensively high computation time and resources. Furthermore, data is often not available due to not only the nature of the problem or privacy concerns but also the cost of data preparation. Data collection, preprocessing, and labeling are strenuous human tasks. Therefore, few-shot learning that could drastically reduce the turnaround time of building machine learning applications emerges as a low-cost solution. This survey paper comprises a representative list of recently proposed few-shot learning algorithms. Given the learning dynamics and characteristics, the approaches to few-shot learning problems are discussed in the perspectives of meta-learning, transfer learning, and hybrid approaches (i.e., different variations of the few-shot learning problem).

cs.LG↗

arXiv2021

Towards Interactive Reinforcement Learning with Intrinsic Feedback

Reinforcement learning (RL) and brain-computer interfaces (BCI) have experienced significant growth over the past decade. With rising interest in human-in-the-loop (HITL), incorporating human input with RL algorithms has given rise to the sub-field of interactive RL. Adjacently, the field of BCI has long been interested in extracting informative brain signals from neural activity for use in human-computer interactions. A key link between these fields lies in the interpretation of neural activity as feedback such that interactive RL approaches can be employed. We denote this new and emerging medium of feedback as intrinsic feedback. Despite intrinsic feedback's ability to be conveyed automatically and even unconsciously, proper exploration surrounding this key link has largely gone unaddressed by both communities. Thus, to help facilitate a deeper understanding and a more effective utilization, we provide a tutorial-style review covering the motivations, approaches, and open problems of intrinsic feedback and its foundational concepts.

cs.AI↗

arXiv2021

Local Learning Matters: Rethinking Data Heterogeneity in Federated Learning

Federated learning (FL) is a promising strategy for performing privacy-preserving, distributed learning with a network of clients (i.e., edge devices). However, the data distribution among clients is often non-IID in nature, making efficient optimization difficult. To alleviate this issue, many FL algorithms focus on mitigating the effects of data heterogeneity across clients by introducing a variety of proximal terms, some incurring considerable compute and/or memory overheads, to restrain local updates with respect to the global model. Instead, we consider rethinking solutions to data heterogeneity in FL with a focus on local learning generality rather than proximal restriction. To this end, we first present a systematic study informed by second-order indicators to better understand algorithm effectiveness in FL. Interestingly, we find that standard regularization methods are surprisingly strong performers in mitigating data heterogeneity effects. Based on our findings, we further propose a simple and effective method, FedAlign, to overcome data heterogeneity and the pitfalls of previous methods. FedAlign achieves competitive accuracy with state-of-the-art FL methods across a variety of settings while minimizing computation and memory overhead. Code is available at https://github.com/mmendiet/FedAlign

cs.LG↗

arXiv2021

Transformation of Node to Knowledge Graph Embeddings for Faster Link Prediction in Social Networks

Recent advances in neural networks have solved common graph problems such as link prediction, node classification, node clustering, node recommendation by developing embeddings of entities and relations into vector spaces. Graph embeddings encode the structural information present in a graph. The encoded embeddings then can be used to predict the missing links in a graph. However, obtaining the optimal embeddings for a graph can be a computationally challenging task specially in an embedded system. Two techniques which we focus on in this work are 1) node embeddings from random walk based methods and 2) knowledge graph embeddings. Random walk based embeddings are computationally inexpensive to obtain but are sub-optimal whereas knowledge graph embeddings perform better but are computationally expensive. In this work, we investigate a transformation model which converts node embeddings obtained from random walk based methods to embeddings obtained from knowledge graph methods directly without an increase in the computational cost. Extensive experimentation shows that the proposed transformation model can be used for solving link prediction in real-time.

cs.SI↗

arXiv2021

Sim-to-Real Transfer in Multi-agent Reinforcement Networking for Federated Edge Computing

Federated Learning (FL) over wireless multi-hop edge computing networks, i.e., multi-hop FL, is a cost-effective distributed on-device deep learning paradigm. This paper presents FedEdge simulator, a high-fidelity Linux-based simulator, which enables fast prototyping, sim-to-real code, and knowledge transfer for multi-hop FL systems. FedEdge simulator is built on top of the hardware-oriented FedEdge experimental framework with a new extension of the realistic physical layer emulator. This emulator exploits trace-based channel modeling and dynamic link scheduling to minimize the reality gap between the simulator and the physical testbed. Our initial experiments demonstrate the high fidelity of the FedEdge simulator and its superior performance on sim-to-real knowledge transfer in reinforcement learning-optimized multi-hop FL.

cs.NI↗

arXiv2021

EdgeML: Towards Network-Accelerated Federated Learning over Wireless Edge

Federated learning (FL) is a distributed machine learning technology for next-generation AI systems that allows a number of workers, i.e., edge devices, collaboratively learn a shared global model while keeping their data locally to prevent privacy leakage. Enabling FL over wireless multi-hop networks can democratize AI and make it accessible in a cost-effective manner. However, the noisy bandwidth-limited multi-hop wireless connections can lead to delayed and nomadic model updates, which significantly slows down the FL convergence speed. To address such challenges, this paper aims to accelerate FL convergence over wireless edge by optimizing the multi-hop federated networking performance. In particular, the FL convergence optimization problem is formulated as a Markov decision process (MDP). To solve such MDP, multi-agent reinforcement learning (MA-RL) algorithms along with domain-specific action space refining schemes are developed, which online learn the delay-minimum forwarding paths to minimize the model exchange latency between the edge devices (i.e., workers) and the remote server. To validate the proposed solutions, FedEdge is developed and implemented, which is the first experimental framework in the literature for FL over multi-hop wireless edge computing networks. FedEdge allows us to fast prototype, deploy, and evaluate novel FL algorithms along with RL-based system optimization methods in real wireless devices. Moreover, a physical experimental testbed is implemented by customizing the widely adopted Linux wireless routers and ML computing nodes.Finally, our experimentation results on the testbed show that the proposed network-accelerated FL system can practically and significantly improve FL convergence speed, compared to the FL system empowered by the production-grade commercially available wireless networking protocol, BATMAN-Adv.

cs.NI↗

arXiv2021

CrossAug: A Contrastive Data Augmentation Method for Debiasing Fact Verification Models

Fact verification datasets are typically constructed using crowdsourcing techniques due to the lack of text sources with veracity labels. However, the crowdsourcing process often produces undesired biases in data that cause models to learn spurious patterns. In this paper, we propose CrossAug, a contrastive data augmentation method for debiasing fact verification models. Specifically, we employ a two-stage augmentation pipeline to generate new claims and evidences from existing samples. The generated samples are then paired cross-wise with the original pair, forming contrastive samples that facilitate the model to rely less on spurious patterns and learn more robust representations. Experimental results show that our method outperforms the previous state-of-the-art debiasing technique by 3.6% on the debiased extension of the FEVER dataset, with a total performance boost of 10.13% from the baseline. Furthermore, we evaluate our approach in data-scarce settings, where models can be more susceptible to biases due to the lack of training data. Experimental results demonstrate that our approach is also effective at debiasing in these low-resource conditions, exceeding the baseline performance on the Symmetric dataset with just 1% of the original data.

cs.CL↗

arXiv2021

An Optimized Dynamic Mode Decomposition Model Robust to Multiplicative Noise

Dynamic mode decomposition (DMD) is an efficient tool for decomposing spatio-temporal data into a set of low-dimensional modes, yielding the oscillation frequencies and the growth rates of physically significant modes. In this paper, we propose a novel DMD model that can be used for dynamical systems affected by multiplicative noise. We first derive a maximum a posteriori (MAP) estimator for the data-based model decomposition of a linear dynamical system corrupted by certain multiplicative noise. Applying penalty relaxation to the MAP estimator, we obtain the proposed DMD model whose epigraphical limits are the MAP estimator and the conventional optimized DMD model. We also propose an efficient alternating gradient descent method for solving the proposed DMD model, and analyze its convergence behavior. The proposed model is demonstrated on both the synthetic data and the numerically generated one-dimensional combustor data, and is shown to have superior reconstruction properties compared to state-of-the-art DMD models. Considering that multiplicative noise is ubiquitous in numerous dynamical systems, the proposed DMD model opens up new possibilities for accurate data-based modal decomposition.

math.DS↗

arXiv2021

MutualNet: Adaptive ConvNet via Mutual Learning from Different Model Configurations

Most existing deep neural networks are static, which means they can only do inference at a fixed complexity. But the resource budget can vary substantially across different devices. Even on a single device, the affordable budget can change with different scenarios, and repeatedly training networks for each required budget would be incredibly expensive. Therefore, in this work, we propose a general method called MutualNet to train a single network that can run at a diverse set of resource constraints. Our method trains a cohort of model configurations with various network widths and input resolutions. This mutual learning scheme not only allows the model to run at different width-resolution configurations but also transfers the unique knowledge among these configurations, helping the model to learn stronger representations overall. MutualNet is a general training methodology that can be applied to various network structures (e.g., 2D networks: MobileNets, ResNet, 3D networks: SlowFast, X3D) and various tasks (e.g., image classification, object detection, segmentation, and action recognition), and is demonstrated to achieve consistent improvements on a variety of datasets. Since we only train the model once, it also greatly reduces the training cost compared to independently training several models. Surprisingly, MutualNet can also be used to significantly boost the performance of a single network, if dynamic resource constraint is not a concern. In summary, MutualNet is a unified method for both static and adaptive, 2D and 3D networks. Codes and pre-trained models are available at \url{https://github.com/taoyang1122/MutualNet}.

cs.CV↗

arXiv2021

System Identification near a Hopf Bifurcation via the Noise-Induced Dynamics in the Fixed-Point Regime

A Hopf bifurcation is prevalent in many nonlinear dynamical systems. When a system prior to a Hopf bifurcation is exposed to a sufficient level of noise, its noise-induced dynamics can provide valuable information about the impending bifurcation. In this thesis, we present a system identification (SI) framework that exploits the noise-induced dynamics prior to a Hopf bifurcation. The framework is novel in that it is capable of predicting the bifurcation point and the post-bifurcation dynamics using only pre-bifurcation data. Specifically, we present two different versions of the framework: input-output and output-only. For the input-output version, the system is forced with additive noise generated by an external actuator, and its response is measured. For the output-only version, the intrinsic noise of the system acts as the noise source and only the output signal is measured. In both versions, the Fokker-Planck equations, which describe the probability density function of the fluctuation amplitude, are derived from self-excited oscillator models. Then, the coefficients of these models are extracted from the experimental probability density functions characterizing the noise-induced response in the fixed-point regime. The SI framework is tested on three different experimental systems: a low-density jet, a flame-driven Rijke tube, and a gas-turbine combustor. For these systems, we demonstrate that the proposed framework can identify the nature of the Hopf bifurcation and the system's order of nonlinearity. Moreover, by extrapolating the identified model coefficients, we are able to forecast the locations of the bifurcation points and the limit-cycle features after those points. To the best of our knowledge, this is the first time that SI has been performed using data from only the pre-bifurcation regime, without the need for knowledge of the location of the bifurcation point.

physics.flu-dyn↗

arXiv2020

Demystifying Deep Neural Networks Through Interpretation: A Survey

Modern deep learning algorithms tend to optimize an objective metric, such as minimize a cross entropy loss on a training dataset, to be able to learn. The problem is that the single metric is an incomplete description of the real world tasks. The single metric cannot explain why the algorithm learn. When an erroneous happens, the lack of interpretability causes a hardness of understanding and fixing the error. Recently, there are works done to tackle the problem of interpretability to provide insights into neural networks behavior and thought process. The works are important to identify potential bias and to ensure algorithm fairness as well as expected performance.

cs.LG↗