arXiv subjects

Da Li

Publications and source records attributed to Da Li.

At least 127 records · Page 7Linked to original sources

arXiv2022

Quasi-static magnetic compression of field-reversed configuration plasma: Amended scalings and limits from two-dimensional MHD equilibrium

In this work, several key scaling laws of the quasi-static magnetic compression of field reversed configuration (FRC) plasma [Spencer, Tuszewski, and Linford, 1983] are amended from a series of 2D FRC MHD equilibriums numerically obtained using the Grad-Shafranov equation solver NIMEQ. Based on the new scaling for the elongation and the magnetic fields at the separatrix and the wall, the empirically stable limits for the compression ratio, the fusion gain, and the neutron yield are evaluated, which may serve as a more accurate estimate for the upper ceiling of performance from the magnetic compression of FRC plasma as a potential fusion energy as well as neutron source devices.

physics.plasm-ph↗

arXiv2022

Pushing the Limits of Simple Pipelines for Few-Shot Learning: External Data and Fine-Tuning Make a Difference

Few-shot learning (FSL) is an important and topical problem in computer vision that has motivated extensive research into numerous methods spanning from sophisticated meta-learning methods to simple transfer learning baselines. We seek to push the limits of a simple-but-effective pipeline for more realistic and practical settings of few-shot image classification. To this end, we explore few-shot learning from the perspective of neural network architecture, as well as a three stage pipeline of network updates under different data supplies, where unsupervised external data is considered for pre-training, base categories are used to simulate few-shot tasks for meta-training, and the scarcely labelled data of an novel task is taken for fine-tuning. We investigate questions such as: (1) How pre-training on external data benefits FSL? (2) How state-of-the-art transformer architectures can be exploited? and (3) How fine-tuning mitigates domain shift? Ultimately, we show that a simple transformer-based pipeline yields surprisingly good performance on standard benchmarks such as Mini-ImageNet, CIFAR-FS, CDFSL and Meta-Dataset. Our code and demo are available at https://hushell.github.io/pmf.

cs.CV↗

arXiv2021

A Channel Coding Benchmark for Meta-Learning

Meta-learning provides a popular and effective family of methods for data-efficient learning of new tasks. However, several important issues in meta-learning have proven hard to study thus far. For example, performance degrades in real-world settings where meta-learners must learn from a wide and potentially multi-modal distribution of training tasks; and when distribution shift exists between meta-train and meta-test task distributions. These issues are typically hard to study since the shape of task distributions, and shift between them are not straightforward to measure or control in standard benchmarks. We propose the channel coding problem as a benchmark for meta-learning. Channel coding is an important practical application where task distributions naturally arise, and fast adaptation to new tasks is practically valuable. We use our MetaCC benchmark to study several aspects of meta-learning, including the impact of task distribution breadth and shift, which can be controlled in the coding problem. Going forward, MetaCC provides a tool for the community to study the capabilities and limitations of meta-learning, and to drive research on practically robust and effective meta-learners.

cs.LG↗

arXiv2021

Weight-Covariance Alignment for Adversarially Robust Neural Networks

Stochastic Neural Networks (SNNs) that inject noise into their hidden layers have recently been shown to achieve strong robustness against adversarial attacks. However, existing SNNs are usually heuristically motivated, and often rely on adversarial training, which is computationally costly. We propose a new SNN that achieves state-of-the-art performance without relying on adversarial training, and enjoys solid theoretical justification. Specifically, while existing SNNs inject learned or hand-tuned isotropic noise, our SNN learns an anisotropic noise distribution to optimize a learning-theoretic bound on adversarial robustness. We evaluate our method on a number of popular benchmarks, show that it can be applied to different architectures, and that it provides robustness to a variety of white-box and black-box attacks, while being simple and fast to train compared to existing alternatives.

cs.LG↗

arXiv2021

Application of an unbalanced optimal transport distance and a mixed L1/Wasserstein distance to full waveform inversion

Full waveform inversion (FWI) is an important and popular technique in subsurface earth property estimation. However, using the least-squares norm in the misfit function often leads to the local minimum solution of the optimization problem, and this phenomenon can be explained with the cycle-skipping artifact. Several methods that apply optimal transport distances to mitigate the cycle-skipping artifact have been proposed recently. The optimal transport distance is designed to compare two probability measures. To overcome the mass equality limit, we introduce an unbalanced optimal transport (UOT) distance with KullbackLeibler divergence to balance the mass difference. Also, a mixed L1/Wasserstein distance is constructed that can preserve the convex properties with respect to shift, dilation, and amplitude change operation. An entropy regularization approach and scaling algorithms are used to compute the distance and the gradient efficiently. Two strategies of normalization methods that transform the seismic signals into non-negative functions are provided. Numerical examples are provided to demonstrate the efficiency and effectiveness of the new method.

math.NA↗

arXiv2021

Angstrom-wide conductive channels in black phosphorus by Cu intercalation

Intercalation is an effective method to improve and modulate properties of two-dimensional materials. Even so, spatially controlled intercalation at atomic scale, which is important to introduce and modulated properties, has not been successful due to difficulties in controlling the diffusion of intercalants. Here, we show formation of angstrom-wide conductive channels (~4.3 A) in black phosphorus by Cu intercalation. The atomic structure, resultant microstructural effects, intercalation mechanism, and local variations of electronic properties modulated in black phosphorus by Cu intercalation were investigated extensively by transmission electron microscopy including in situ observation, DFT calculation, and conductive atomic force microscopy.

cond-mat.mtrl-sci↗

arXiv2020

Online Meta-Learning for Multi-Source and Semi-Supervised Domain Adaptation

Domain adaptation (DA) is the topical problem of adapting models from labelled source datasets so that they perform well on target datasets where only unlabelled or partially labelled data is available. Many methods have been proposed to address this problem through different ways to minimise the domain shift between source and target datasets. In this paper we take an orthogonal perspective and propose a framework to further enhance performance by meta-learning the initial conditions of existing DA algorithms. This is challenging compared to the more widely considered setting of few-shot meta-learning, due to the length of the computation graph involved. Therefore we propose an online shortest-path meta-learning framework that is both computationally tractable and practically effective for improving DA performance. We present variants for both multi-source unsupervised domain adaptation (MSDA), and semi-supervised domain adaptation (SSDA). Importantly, our approach is agnostic to the base adaptation algorithm, and can be applied to improve many techniques. Experimentally, we demonstrate improvements on classic (DANN) and recent (MCD and MME) techniques for MSDA and SSDA, and ultimately achieve state of the art results on several DA benchmarks including the largest scale DomainNet.

cs.CV↗

arXiv2020

Sequential Learning for Domain Generalization

In this paper we propose a sequential learning framework for Domain Generalization (DG), the problem of training a model that is robust to domain shift by design. Various DG approaches have been proposed with different motivating intuitions, but they typically optimize for a single step of domain generalization -- training on one set of domains and generalizing to one other. Our sequential learning is inspired by the idea lifelong learning, where accumulated experience means that learning the $n^{th}$ thing becomes easier than the $1^{st}$ thing. In DG this means encountering a sequence of domains and at each step training to maximise performance on the next domain. The performance at domain $n$ then depends on the previous $n-1$ learning problems. Thus backpropagating through the sequence means optimizing performance not just for the next domain, but all following domains. Training on all such sequences of domains provides dramatically more `practice' for a base DG learner compared to existing approaches, thus improving performance on a true testing domain. This strategy can be instantiated for different base DG algorithms, but we focus on its application to the recently proposed Meta-Learning Domain generalization (MLDG). We show that for MLDG it leads to a simple to implement and fast algorithm that provides consistent performance improvement on a variety of DG benchmarks.

cs.CV↗

arXiv2020

Efficient and Stable Finite Difference Modelling of Acoustic Wave Propagation in Variable-density Media

In this paper, we consider the development and analysis of a new explicit compact high-order finite difference scheme for acoustic wave equation formulated in divergence form, which is widely used to describe seismic wave propagation through a heterogeneous media with variable media density and acoustic velocity. The new scheme is compact and of fourth-order accuracy in space and second-order accuracy in time. The compactness of the scheme is obtained by the so-called combined finite difference method, which utilizes the boundary values of the spatial derivatives and those boundary values are obtained by one-sided finite difference approximation. An empirical stability analysis has been conducted to obtain the Currant-Friedrichs-Lewy (CFL) condition, which confirmed the conditional stability of the new scheme. Four numerical examples have been conducted to validate the convergence and effectiveness of the new scheme. The application of the new scheme to a realistic wave propagation problem with Perfect Matched Layer boundary condition is also validated in this paper as well.

math.NA↗

arXiv2019

Episodic Training for Domain Generalization

Domain generalization (DG) is the challenging and topical problem of learning models that generalize to novel testing domains with different statistics than a set of known training domains. The simple approach of aggregating data from all source domains and training a single deep neural network end-to-end on all the data provides a surprisingly strong baseline that surpasses many prior published methods. In this paper, we build on this strong baseline by designing an episodic training procedure that trains a single deep network in a way that exposes it to the domain shift that characterises a novel domain at runtime. Specifically, we decompose a deep network into feature extractor and classifier components, and then train each component by simulating it interacting with a partner who is badly tuned for the current domain. This makes both components more robust, ultimately leading to our networks producing state-of-the-art performance on three DG benchmarks. Furthermore, we consider the pervasive workflow of using an ImageNet trained CNN as a fixed feature extractor for downstream recognition tasks. Using the Visual Decathlon benchmark, we demonstrate that our episodic-DG training improves the performance of such a general-purpose feature extractor by explicitly training a feature for robustness to novel problems. This shows that DG training can benefit standard practice in computer vision.

cs.CV↗

arXiv2019

Room temperature 2D ferromagnetism in few-layered 1$T$-CrTe$_{2}$

Spin-related electronics using two dimensional (2D) van der Waals (vdW) materials as a platform are believed to hold great promise for revolutionizing the next generation spintronics. Although many emerging new phenomena have been unravelled in 2D electronic systems with spin long-range orderings, the scarcely reported room temperature magnetic vdW material has thus far hindered the related applications. Here, we show that intrinsic ferromagnetically aligned spin polarization can hold up to 316 K in a metallic phase of 1$T$-CrTe$_{2}$ in the few-layer limit. This room temperature 2D long range spin interaction may be beneficial from an itinerant enhancement. Spin transport measurements indicate an in-plane room temperature negative anisotropic magnetoresistance (AMR) in few-layered CrTe$_{2}$, but a sign change in the AMR at lower temperature, with -0.6$\%$ at 300 K and +5$\%$ at 10 K, respectively. This behavior may originate from the specific spin polarized band structure of CrTe$_{2}$. Our findings provide insights into magnetism in few-layered CrTe$_{2}$, suggesting potential for future room temperature spintronic applications of such 2D vdW magnets.

cond-mat.mes-hall↗

arXiv2019

Large-Scale Pedestrian Retrieval Competition

The Large-Scale Pedestrian Retrieval Competition (LSPRC) mainly focuses on person retrieval which is an important end application in intelligent vision system of surveillance. Person retrieval aims at searching the interested target with specific visual attributes or images. The low image quality, various camera viewpoints, large pose variations and occlusions in real scenes make it a challenge problem. By providing large-scale surveillance data in real scene and standard evaluation methods that are closer to real application, the competition aims to improve the robust of related algorithms and further meet the complicated situations in real application. LSPRC includes two kinds of tasks, i.e., Attribute based Pedestrian Retrieval (PR-A) and Re-IDentification (ReID) based Pedestrian Retrieval (PR-ID). The normal evaluation index, i.e., mean Average Precision (mAP), is used to measure the performances of the two tasks under various scale, pose and occlusion. While the method of system evaluation is introduced to evaluate the person retrieval system in which the related algorithms of the two tasks are integrated into a large-scale video parsing platform (named ISEE) combing with algorithm of pedestrian detection.

cs.CV↗

arXiv2018

Deep Factorised Inverse-Sketching

Modelling human free-hand sketches has become topical recently, driven by practical applications such as fine-grained sketch based image retrieval (FG-SBIR). Sketches are clearly related to photo edge-maps, but a human free-hand sketch of a photo is not simply a clean rendering of that photo's edge map. Instead there is a fundamental process of abstraction and iconic rendering, where overall geometry is warped and salient details are selectively included. In this paper we study this sketching process and attempt to invert it. We model this inversion by translating iconic free-hand sketches to contours that resemble more geometrically realistic projections of object boundaries, and separately factorise out the salient added details. This factorised re-representation makes it easier to match a free-hand sketch to a photo instance of an object. Specifically, we propose a novel unsupervised image style transfer model based on enforcing a cyclic embedding consistency constraint. A deep FG-SBIR model is then formulated to accommodate complementary discriminative detail from each factorised sketch for better matching with the corresponding photo. Our method is evaluated both qualitatively and quantitatively to demonstrate its superiority over a number of state-of-the-art alternatives for style transfer and FG-SBIR.

cs.CV↗

arXiv2018

Broken mirror symmetry tuned topological transport in PbTe/SnTe heterostructures

The tunability of topological surface states and controllable opening of the Dirac gap are of great importance to the application of topological materials. In topological crystalline insulators (TCIs), crystal symmetry and topology of electronic bands intertwine to create topological surface states and thus the Dirac gap can be modulated by symmetry breaking structural changes of lattice. By transport measurement on heterostructures composed of p-type topological crystalline insulator SnTe and n-type conventional semiconductor PbTe, here we show a giant linear magnetoresistance (up to 2150% under 14 T at 2 K) induced by the Dirac Fermions at the PbTe/SnTe interface. In contrast, PbTe/SnTe samples grown at elevated temperature exhibit a cubic-to-rhombohedral structural phase transition of SnTe lattice below 100 K and weak antilocalization effect. Such distinctive magneto-resistance behavior is attributed to the broken mirror symmetry and gapping of topological surface states. Our work provides a promising application for future magneto-electronics and spintronics based on TCI heterostructures.

cond-mat.mtrl-sci↗

arXiv2018

Sketch-a-Classifier: Sketch-based Photo Classifier Generation

Contemporary deep learning techniques have made image recognition a reasonably reliable technology. However training effective photo classifiers typically takes numerous examples which limits image recognition's scalability and applicability to scenarios where images may not be available. This has motivated investigation into zero-shot learning, which addresses the issue via knowledge transfer from other modalities such as text. In this paper we investigate an alternative approach of synthesizing image classifiers: almost directly from a user's imagination, via free-hand sketch. This approach doesn't require the category to be nameable or describable via attributes as per zero-shot learning. We achieve this via training a {model regression} network to map from {free-hand sketch} space to the space of photo classifiers. It turns out that this mapping can be learned in a category-agnostic way, allowing photo classifiers for new categories to be synthesized by user with no need for annotated training photos. {We also demonstrate that this modality of classifier generation can also be used to enhance the granularity of an existing photo classifier, or as a complement to name-based zero-shot learning.

cs.CV↗

arXiv2018

Chromium-Induced Ferromagnetism with Perpendicular Anisotropy in Topological Crystalline Insulator SnTe (111) Thin Films

Topological crystalline insulator (TCI) is a recently-discovered topological phase of matter. It possesses multiple Dirac surface states, which are protected by the crystal symmetry. This is in contrast to the time reversal symmetry that is operative in the well-known topological insulators. In the presence of a Zeeman field and/or strain, the multiple Dirac surface states are gapped. The high-Chern-number quantum anomalous Hall (QAH) state is predicted to emerge if the chemical potential resides in all the Zeeman gaps. Here, we use molecular beam epitaxy to grow 12 double layer (DL) pure and Cr-doped SnTe (111) thin film on heat-treated SrTiO3 (111) substrate using a quintuple layer of insulating (Bi0.2Sb0.8)2Te3 topological insulator as a buffer film. The Hall traces of Cr-doped SnTe film at low temperatures display square hysteresis loops indicating long-range ferromagnetic order with perpendicular anisotropy. The Curie temperature of the 12DL Sn0.9Cr0.1Te film is ~ 110 K. Due to the chemical potential crossing the bulk valence bands, the anomalous Hall resistance of 12DL Sn0.9Cr0.1Te film is substantially lower than the predicted quantized value (~1/4 h/e2). It is possible that with systematic tuning the chemical potential via chemical doping and electrical gating, the high-Chern-number QAH state can be realized in the Cr-doped SnTe (111) thin film.

cond-mat.mes-hall↗

arXiv2018

Electric-field control of magnetism in few-layered van der Waals magnet

Manipulating quantum state via electrostatic gating has been intriguing for many model systems in nanoelectronics. When it comes to the question of controlling the electron spins, more specifically, the magnetism of a system, tuning with electric field has been proven to be elusive. Recently, magnetic layered semiconductors have attracted much attention due to their emerging new physical phenomena. However, challenges still remain in the demonstration of a gate controllable magnetism based on them. Here, we show that, via ionic gating, strong field effect can be observed in few-layered semiconducting Cr$_{2}$Ge$_{2}$Te$_{6}$ devices. At different gate doping, micro-area Kerr measurements in the studied devices demonstrate tunable magnetization loops below the Curie temperature, which is tentatively attributed to the moment re-balance in the spin-polarized band structure. Our findings of electric-field controlled magnetism in van der Waals magnets pave the way for potential applications in new generation magnetic memory storage, sensors, and spintronics.

cond-mat.mes-hall↗

arXiv2017

Fast Integral Histogram Computations on GPU for Real-Time Video Analytics

In many Multimedia content analytics frameworks feature likelihood maps represented as histograms play a critical role in the overall algorithm. Integral histograms provide an efficient computational framework for extracting multi-scale histogram-based regional descriptors in constant time which are considered as the principle building blocks of many video content analytics frameworks. We evaluate four different mappings of the integral histogram computation onto Graphics Processing Units (GPUs) using different kernel optimization strategies. Our kernels perform cumulative sums on row and column histograms in a cross-weave or wavefront scan order, use different data organization and scheduling methods that is shown to critically affect utilization of GPU resources (cores and shared memory). Tiling the 3-D array into smaller regular data blocks significantly speeds up the efficiency of the computation compared to a strip-based organization. The tiled integral histogram using a diagonal wavefront scan has the best performance of about 300.4 frames/sec for 640 x 480 images and 32 bins with a speedup factor of about 120 using GTX Titan X graphics card compared to a single threaded sequential CPU implementation. Double-buffering has been exploited to overlap computation and communication across sequence of images. Mapping integral histogram bins computations onto multiple GPUs enables us to process 32 giga bytes integral histogram data (of 64MB Image and 128 bins) with a frame rate of 0.73 Hz and speedup factor of 153X over single-threaded CPU implementation and the speedup of 45X over 16-threaded CPU implementation.

cs.DC↗