Searcharxiv⌕ Search

arXiv subjects

Ning Xu

Publications and source records attributed to Ning Xu.

At least 73 records · Page 4Linked to original sources
arXiv2023

Unreliable Partial Label Learning with Recursive Separation

Partial label learning (PLL) is a typical weakly supervised learning problem in which each instance is associated with a candidate label set, and among which only one is true. However, the assumption that the ground-truth label is always among the candidate label set would be unrealistic, as the reliability of the candidate label sets in real-world applications cannot be guaranteed by annotators. Therefore, a generalized PLL named Unreliable Partial Label Learning (UPLL) is proposed, in which the true label may not be in the candidate label set. Due to the challenges posed by unreliable labeling, previous PLL methods will experience a marked decline in performance when applied to UPLL. To address the issue, we propose a two-stage framework named Unreliable Partial Label Learning with Recursive Separation (UPLLRS). In the first stage, the self-adaptive recursive separation strategy is proposed to separate the training set into a reliable subset and an unreliable subset. In the second stage, a disambiguation strategy is employed to progressively identify the ground-truth labels in the reliable subset. Simultaneously, semi-supervised learning methods are adopted to extract valuable information from the unreliable subset. Our method demonstrates state-of-the-art performance as evidenced by experimental results, particularly in situations of high unreliability. Code and supplementary materials are available at https://github.com/dhiyu/UPLLRS.

cs.AI↗
arXiv2023

Variational Label-Correlation Enhancement for Congestion Prediction

The physical design process of large-scale designs is a time-consuming task, often requiring hours to days to complete, with routing being the most critical and complex step. As the the complexity of Integrated Circuits (ICs) increases, there is an increased demand for accurate routing quality prediction. Accurate congestion prediction aids in identifying design flaws early on, thereby accelerating circuit design and conserving resources. Despite the advancements in current congestion prediction methodologies, an essential aspect that has been largely overlooked is the spatial label-correlation between different grids in congestion prediction. The spatial label-correlation is a fundamental characteristic of circuit design, where the congestion status of a grid is not isolated but inherently influenced by the conditions of its neighboring grids. In order to fully exploit the inherent spatial label-correlation between neighboring grids, we propose a novel approach, {\ours}, i.e., VAriational Label-Correlation Enhancement for Congestion Prediction, which considers the local label-correlation in the congestion map, associating the estimated congestion value of each grid with a local label-correlation weight influenced by its surrounding grids. {\ours} leverages variational inference techniques to estimate this weight, thereby enhancing the regression model's performance by incorporating spatial dependencies. Experiment results validate the superior effectiveness of {\ours} on the public available \texttt{ISPD2011} and \texttt{DAC2012} benchmarks using the superblue circuit line.

cs.LG↗
arXiv2023

Oxidation kinetics and non-Marcusian charge transfer in dimensionally confined semiconductors

Electrochemical reactions represent essential processes in fundamental chemistry that foster a wide range of applications. Although most electrochemical reactions in bulk substances can be well described by the classical Marcus-Gerischer charge transfer theory, the realistic reaction character and mechanism in dimensionally confined systems remain unknown. Here, we report the multiparametric survey on the kinetics of lateral photooxidation in structurally identical WS2 and MoS2 monolayers, where electrochemical oxidation occurs at the atomically thin monolayer edges. The oxidation rate is correlated quantitatively with various crystallographic and environmental parameters, including the density of reactive sites, humidity, temperature, and illumination fluence. In particular, we observe distinctive reaction barriers of 1.4 and 0.9 eV for the two structurally identical semiconductors and uncover an unusual non-Marcusian charge transfer mechanism in these dimensionally confined monolayers due to the limit in reactant supplies. A scenario of band bending is proposed to explain the discrepancy in reaction barriers. These results add important knowledge into the fundamental electrochemical reaction theory in low-dimensional systems.

cond-mat.mtrl-sci↗
arXiv2023

Human in Events: A Large-Scale Benchmark for Human-centric Video Analysis in Complex Events

Along with the development of modern smart cities, human-centric video analysis has been encountering the challenge of analyzing diverse and complex events in real scenes. A complex event relates to dense crowds, anomalous individuals, or collective behaviors. However, limited by the scale and coverage of existing video datasets, few human analysis approaches have reported their performances on such complex events. To this end, we present a new large-scale dataset with comprehensive annotations, named Human-in-Events or HiEve (Human-centric video analysis in complex Events), for the understanding of human motions, poses, and actions in a variety of realistic events, especially in crowd & complex events. It contains a record number of poses (>1M), the largest number of action instances (>56k) under complex events, as well as one of the largest numbers of trajectories lasting for longer time (with an average trajectory length of >480 frames). Based on its diverse annotation, we present two simple baselines for action recognition and pose estimation, respectively. They leverage cross-label information during training to enhance the feature learning in corresponding visual tasks. Experiments show that they could boost the performance of existing action recognition and pose estimation pipelines. More importantly, they prove the widely ranged annotations in HiEve can improve various video tasks. Furthermore, we conduct extensive experiments to benchmark recent video analysis approaches together with our baseline methods, demonstrating HiEve is a challenging dataset for human-centric video analysis. We expect that the dataset will advance the development of cutting-edge techniques in human-centric analysis and the understanding of complex events. The dataset is available at http://humaninevents.org

cs.CV↗
arXiv2023

Unraveling Intertwined Impacts between Lattice Vacancy and Substrate on Photonic Quasiparticles in Monolayer MoS$_2$

Lattice defects and interfacial absorbates represent two extrinsic but ubiquitous factors that exert profound impacts on the luminescent properties of semiconductors. However, their impacts are normally tangled and remain to be separately elucidated. Here, we clarify the individual roles of each factor by tracking the weight evolution of photonic quasiparticles by modulating the densities of sulfur vacancies in monolayer MoS$_2$ via ad hoc defect engineering. In particular, we perform atomistic analyses on the densities of sulfur vacancies by employing atomically resolved scanning transmission electron microscopy to quantitatively characterize the generation rates of sulfur vacancies in MoS$_2$ on two types of substrates with opposite surface hydrophobicity natures. We reveal that the generation rate of sulfur vacancies can even double on the hydrophilic substrates. More importantly, the impact of sulfur vacancies over the weight of trionic emission is closely associated with the substrate hydrophobicity, which is manifested noticeably in the hydrophobic substrates, but insignificant in the hydrophilic ones. The results represent an in-depth understanding on the roles of the extrinsic factors on the luminescent properties in atomically thin semiconductors.

cond-mat.mtrl-sci↗
arXiv2023

Intrinsic Electronic Properties of BN Encapsulated, van der Waals Contacted MoSe$_2$ FETs

Two-dimensional (2D) semiconductors have attracted considerable interest for their unique physical properties. Here, we report the intrinsic cryogenic electronic transport properties in few-layer MoSe$_2$ field-effect transistors (FETs) that are simultaneously van der Waals contacted with gold electrodes and are fully encapsulated in ultraclean hexagonal boron nitride dielectrics. The FETs exhibit electronically favorable channel/dielectric interfaces with low densities of interfacial traps ($10^{10}\,$cm$^{-2}$), which lead to outstanding device characteristics at room temperature, including a near-Boltzmann-limit subthreshold swings ($\lt65\,$mV/dec), a high carrier mobility ($68\,$cm$^{2}\cdot$V$^{-1}\cdot$s$^{-1}$), and a negligible scanning hysteresis ($\lt15\,$mV). The dependence of various contact-related quantities on temperature and carrier density are also systematically characterized to understand the van der Waals contacts between gold and MoSe$_2$. The results provide insightful information on the device physics in van der Waals contacted and encapsulated 2D FETs.

cond-mat.mtrl-sci↗
arXiv2023

Noninvasive Photodelamination of van der Waals Semiconductors for High-Performance Electronics

Atomically thin two-dimensional (2D) van der Waals semiconductors are promising candidate materials for post-silicon electronics. However, it remains challenging to attain completely uniform monolayer semiconductor wafers free of over-grown islands. Here, we report the observation of the energy funneling effect and ambient photodelamination phenomenon in inhomogeneous few-layer WS$_2$ flakes under low illumination fluencies down to several nW/$μ$m$^{2}$ and its potential as a non-invasive post-etching strategy for selectively stripping the local excessive overlying islands. Photoluminescent tracking on the photoetching traces reveals relatively fast etching rates around $0.3-0.8\,μ$m/min at varied temperatures and an activation energy of $1.7\,$eV. By using crystallographic and electronic characterization, we also confirm the non-invasive nature of the low-power photodelamination and the highly preserved lattice quality in the as-etched monolayer products, featuring a comparable average density of atomic defects (ca.$4.2\times 10^{13}\,$cm$^{-2}$) to pristine flakes and a high electron mobility up to $80\,$cm$^{2}\cdot$V$^{-1}\cdot$s$^{-1}$) at room temperature. This approach opens a non-invasive photoetching route for thickness uniformity management in 2D van der Waals semiconductor wafers for electronic applications.

cond-mat.mtrl-sci↗
arXiv2023

Dynamic Gardner crossover in a simple structural glass

The criticality of the jamming transition responsible for amorphous solidification has been theoretically linked to the marginal stability of a thermodynamic Gardner phase. While the critical exponents of jamming appear independent of the preparation history, the pertinence of Gardner physics far from equilibrium is an open question. To fill this gap, we numerically study the nonequilibrium dynamics of hard disks compressed towards the jamming transition using a broad variety of protocols. We show that dynamic signatures of Gardner physics can be disentangled from the aging relaxation dynamics. We thus define a generic dynamic Gardner crossover regardless of the history. Our results show that the jamming transition is always accessed by exploring increasingly complex landscape, resulting in the anomalous microscopic relaxation dynamics that remains to be understood theoretically.

cond-mat.dis-nn↗
arXiv2023

Progressive Purification for Instance-Dependent Partial Label Learning

Partial label learning (PLL) aims to train multiclass classifiers from the examples each annotated with a set of candidate labels where a fixed but unknown candidate label is correct. In the last few years, the instance-independent generation process of candidate labels has been extensively studied, on the basis of which many theoretical advances have been made in PLL. Nevertheless, the candidate labels are always instance-dependent in practice and there is no theoretical guarantee that the model trained on the instance-dependent PLL examples can converge to an ideal one. In this paper, a theoretically grounded and practically effective approach named POP, i.e. PrOgressive Purification for instance-dependent partial label learning, is proposed. Specifically, POP updates the learning model and purifies each candidate label set progressively in every epoch. Theoretically, we prove that POP enlarges the region appropriately fast where the model is reliable, and eventually approximates the Bayes optimal classifier with mild assumptions. Technically, POP is flexible with arbitrary PLL losses and could improve the performance of the previous PLL losses in the instance-dependent case. Experiments on the benchmark datasets and the real-world datasets validate the effectiveness of the proposed method.

cs.LG↗
arXiv2023

Towards Effective Visual Representations for Partial-Label Learning

Under partial-label learning (PLL) where, for each training instance, only a set of ambiguous candidate labels containing the unknown true label is accessible, contrastive learning has recently boosted the performance of PLL on vision tasks, attributed to representations learned by contrasting the same/different classes of entities. Without access to true labels, positive points are predicted using pseudo-labels that are inherently noisy, and negative points often require large batches or momentum encoders, resulting in unreliable similarity information and a high computational overhead. In this paper, we rethink a state-of-the-art contrastive PLL method PiCO[24], inspiring the design of a simple framework termed PaPi (Partial-label learning with a guided Prototypical classifier), which demonstrates significant scope for improvement in representation learning, thus contributing to label disambiguation. PaPi guides the optimization of a prototypical classifier by a linear classifier with which they share the same feature encoder, thus explicitly encouraging the representation to reflect visual similarity between categories. It is also technically appealing, as PaPi requires only a few components in PiCO with the opposite direction of guidance, and directly eliminates the contrastive learning module that would introduce noise and consume computational resources. We empirically demonstrate that PaPi significantly outperforms other PLL methods on various image classification tasks.

cs.CV↗
arXiv2023

A Distribution Evolutionary Algorithm for the Graph Coloring Problem

Graph coloring is a challenging combinatorial optimization problem with a wide range of applications. In this paper, a distribution evolutionary algorithm based on a population of probability model (DEA-PPM) is developed to address it efficiently. Unlike existing estimation of distribution algorithms where a probability model is updated by generated solutions, DEA-PPM employs a distribution population based on a novel probability model, and an orthogonal exploration strategy is introduced to search the distribution space with the assistance of an refinement strategy. By sampling the distribution population, efficient search in the solution space is realized based on a tabu search process. Meanwhile, DEA-PPM introduces an iterative vertex removal strategy to improve the efficiency of $k$-coloring, and an inherited initialization strategy is implemented to address the chromatic problem well. The cooperative evolution of the distribution population and the solution population leads to a good balance between exploration and exploitation. Numerical results demonstrate that the DEA-PPM of small population size is competitive to the state-of-the-art metaheuristics.utes to its competitiveness to the state-of-the-art metaheuristics.

cs.NE↗
arXiv2023

Adaptive Cross-Layer Attention for Image Restoration

Non-local attention module has been proven to be crucial for image restoration. Conventional non-local attention processes features of each layer separately, so it risks missing correlation between features among different layers. To address this problem, we aim to design attention modules that aggregate information from different layers. Instead of finding correlated key pixels within the same layer, each query pixel is encouraged to attend to key pixels at multiple previous layers of the network. In order to efficiently embed such attention design into neural network backbones, we propose a novel Adaptive Cross-Layer Attention (ACLA) module. Two adaptive designs are proposed for ACLA: (1) adaptively selecting the keys for non-local attention at each layer; (2) automatically searching for the insertion locations for ACLA modules. By these two adaptive designs, ACLA dynamically selects a flexible number of keys to be aggregated for non-local attention at previous layer while maintaining a compact neural network with compelling performance. Extensive experiments on image restoration tasks, including single image super-resolution, image denoising, image demosaicing, and image compression artifacts reduction, validate the effectiveness and efficiency of ACLA. The code of ACLA is available at \url{https://github.com/SDL-ASU/ACLA}.

eess.IV↗
arXiv2023

Scalable Multiple Patterning Layout Decomposition Implemented by a Distribution Evolutionary Algorithm

As the feature size of semiconductor technology shrinks to 10 nm and beyond, the multiple patterning lithography (MPL) attracts more attention from the industry. In this paper, we model the layout decomposition of MPL as a generalized graph coloring problem, which is addressed by a distribution evolutionary algorithm based on a population of probabilistic model (DEA-PPM). DEA-PPM can strike a balance between decomposition results and running time, being scalable for varied settings of mask number and lithography resolution. Due to its robustness of decomposition results, this could be an alternative technique for multiple patterning layout decomposition in next-generation technology nodes.

cs.NE↗
arXiv2023

Mechanical-scan-free and multi-color super-resolution imaging with diffractive spot array illumination

Point-scanning microscopy approaches are transforming super-resolution imaging. Despite achieving parallel high-speed imaging using multifocal techniques, efficient multi-color capability with high-quality illumination is currently lacking. In this paper, we present for the first time Mechanical-scan-free and multi-Color Super-resolution Microscopy (MCoSM) by spot array illumination, which enables mechanical-scan-free super-resolution imaging with adjustable resolution and field of view (FoV) based on spatial light modulators (SLMs). Through 100s-10,000s super-resolution spot illumination with different FoV for imaging, we demonstrate the adjustable capacity of MCoSM. MCoSM extends current spectral imaging capabilities through a time-sharing process of different color illumination with phase-shifting scanning, while retaining the spatial flexibility of super-resolution imaging with diffractive spot array illumination. To showcase the prospects for further combining MCoSM with multi-color imaging, we also perform spectral unmixing (four-colors) on images of fluorescent beads at high resolution. MCoSM provides a versatile platform for studying molecular interactions in complex samples at the nanoscale level.

physics.optics↗
arXiv2023

Learning From Biased Soft Labels

Knowledge distillation has been widely adopted in a variety of tasks and has achieved remarkable successes. Since its inception, many researchers have been intrigued by the dark knowledge hidden in the outputs of the teacher model. Recently, a study has demonstrated that knowledge distillation and label smoothing can be unified as learning from soft labels. Consequently, how to measure the effectiveness of the soft labels becomes an important question. Most existing theories have stringent constraints on the teacher model or data distribution, and many assumptions imply that the soft labels are close to the ground-truth labels. This paper studies whether biased soft labels are still effective. We present two more comprehensive indicators to measure the effectiveness of such soft labels. Based on the two indicators, we give sufficient conditions to ensure biased soft label based learners are classifier-consistent and ERM learnable. The theory is applied to three weakly-supervised frameworks. Experimental results validate that biased soft labels can also teach good students, which corroborates the soundness of the theory.

cs.LG↗
arXiv2023

Decompositional Generation Process for Instance-Dependent Partial Label Learning

Partial label learning (PLL) is a typical weakly supervised learning problem, where each training example is associated with a set of candidate labels among which only one is true. Most existing PLL approaches assume that the incorrect labels in each training example are randomly picked as the candidate labels and model the generation process of the candidate labels in a simple way. However, these approaches usually do not perform as well as expected due to the fact that the generation process of the candidate labels is always instance-dependent. Therefore, it deserves to be modeled in a refined way. In this paper, we consider instance-dependent PLL and assume that the generation process of the candidate labels could decompose into two sequential parts, where the correct label emerges first in the mind of the annotator but then the incorrect labels related to the feature are also selected with the correct label as candidate labels due to uncertainty of labeling. Motivated by this consideration, we propose a novel PLL method that performs Maximum A Posterior (MAP) based on an explicitly modeled generation process of candidate labels via decomposed probability distribution models. Extensive experiments on manually corrupted benchmark datasets and real-world datasets validate the effectiveness of the proposed method. Source code is available at https://github.com/palm-ml/idgp.

cs.LG↗
arXiv2022

On the Robustness of Average Losses for Partial-Label Learning

Partial-label learning (PLL) utilizes instances with PLs, where a PL includes several candidate labels but only one is the true label (TL). In PLL, identification-based strategy (IBS) purifies each PL on the fly to select the (most likely) TL for training; average-based strategy (ABS) treats all candidate labels equally for training and let trained models be able to predict TL. Although PLL research has focused on IBS for better performance, ABS is also worthy of study since modern IBS behaves like ABS in the beginning of training to prepare for PL purification and TL selection. In this paper, we analyze why ABS was unsatisfactory and propose how to improve it. Theoretically, we formalize five problem settings of PLL and prove that average PL losses (APLLs) with bounded multi-class losses are always robust, while APLLs with unbounded losses may be non-robust, which is the first robustness analysis for PLL. Experimentally, we have two promising findings: ABS using bounded losses can match/exceed state-of-the-art performance of IBS using unbounded losses; after using robust APLLs to warm start, IBS can further improve upon itself. Our work draws attention to ABS research, which can in turn boost IBS and push forward the whole PLL.

cs.LG↗
arXiv2022

One Positive Label is Sufficient: Single-Positive Multi-Label Learning with Label Enhancement

Multi-label learning (MLL) learns from the examples each associated with multiple labels simultaneously, where the high cost of annotating all relevant labels for each training example is challenging for real-world applications. To cope with the challenge, we investigate single-positive multi-label learning (SPMLL) where each example is annotated with only one relevant label, and show that one can successfully learn a theoretically grounded multi-label classifier for the problem. In this paper, a novel SPMLL method named SMILE, i.e., Single-positive MultI-label learning with Label Enhancement, is proposed. Specifically, an unbiased risk estimator is derived, which could be guaranteed to approximately converge to the optimal risk minimizer of fully supervised learning and shows that one positive label of each instance is sufficient to train the predictive model. Then, the corresponding empirical risk estimator is established via recovering the latent soft label as a label enhancement process, where the posterior density of the latent soft labels is approximate to the variational Beta density parameterized by an inference model. Experiments on benchmark datasets validate the effectiveness of the proposed method.

cs.LG↗