Searcharxiv⌕ Search

arXiv subjects

Can Wang

Publications and source records attributed to Can Wang.

At least 145 records · Page 8Linked to original sources
arXiv2024

Knowledge Translation: A New Pathway for Model Compression

Deep learning has witnessed significant advancements in recent years at the cost of increasing training, inference, and model storage overhead. While existing model compression methods strive to reduce the number of model parameters while maintaining high accuracy, they inevitably necessitate the re-training of the compressed model or impose architectural constraints. To overcome these limitations, this paper presents a novel framework, termed \textbf{K}nowledge \textbf{T}ranslation (KT), wherein a ``translation'' model is trained to receive the parameters of a larger model and generate compressed parameters. The concept of KT draws inspiration from language translation, which effectively employs neural networks to convert different languages, maintaining identical meaning. Accordingly, we explore the potential of neural networks to convert models of disparate sizes, while preserving their functionality. We propose a comprehensive framework for KT, introduce data augmentation strategies to enhance model performance despite restricted training data, and successfully demonstrate the feasibility of KT on the MNIST dataset. Code is available at \url{https://github.com/zju-SWJ/KT}.

cs.LG↗
arXiv2023

OpenGSL: A Comprehensive Benchmark for Graph Structure Learning

Graph Neural Networks (GNNs) have emerged as the de facto standard for representation learning on graphs, owing to their ability to effectively integrate graph topology and node attributes. However, the inherent suboptimal nature of node connections, resulting from the complex and contingent formation process of graphs, presents significant challenges in modeling them effectively. To tackle this issue, Graph Structure Learning (GSL), a family of data-centric learning approaches, has garnered substantial attention in recent years. The core concept behind GSL is to jointly optimize the graph structure and the corresponding GNN models. Despite the proposal of numerous GSL methods, the progress in this field remains unclear due to inconsistent experimental protocols, including variations in datasets, data processing techniques, and splitting strategies. In this paper, we introduce OpenGSL, the first comprehensive benchmark for GSL, aimed at addressing this gap. OpenGSL enables a fair comparison among state-of-the-art GSL methods by evaluating them across various popular datasets using uniform data processing and splitting strategies. Through extensive experiments, we observe that existing GSL methods do not consistently outperform vanilla GNN counterparts. We also find that there is no significant correlation between the homophily of the learned structure and task performance, challenging the common belief. Moreover, we observe that the learned graph structure demonstrates a strong generalization ability across different GNN models, despite the high computational and space consumption. We hope that our open-sourced library will facilitate rapid and equitable evaluation and inspire further innovative research in this field. The code of the benchmark can be found in https://github.com/OpenGSL/OpenGSL.

cs.LG↗
arXiv2023

StyleRetoucher: Generalized Portrait Image Retouching with GAN Priors

Creating fine-retouched portrait images is tedious and time-consuming even for professional artists. There exist automatic retouching methods, but they either suffer from over-smoothing artifacts or lack generalization ability. To address such issues, we present StyleRetoucher, a novel automatic portrait image retouching framework, leveraging StyleGAN's generation and generalization ability to improve an input portrait image's skin condition while preserving its facial details. Harnessing the priors of pretrained StyleGAN, our method shows superior robustness: a). performing stably with fewer training samples and b). generalizing well on the out-domain data. Moreover, by blending the spatial features of the input image and intermediate features of the StyleGAN layers, our method preserves the input characteristics to the largest extent. We further propose a novel blemish-aware feature selection mechanism to effectively identify and remove the skin blemishes, improving the image skin condition. Qualitative and quantitative evaluations validate the great generalization capability of our method. Further experiments show StyleRetoucher's superior performance to the alternative solutions in the image retouching task. We also conduct a user perceptive study to confirm the superior retouching performance of our method over the existing state-of-the-art alternatives.

cs.CV↗
arXiv2023

Mesh-Guided Neural Implicit Field Editing

Neural implicit fields have emerged as a powerful 3D representation for reconstructing and rendering photo-realistic views, yet they possess limited editability. Conversely, explicit 3D representations, such as polygonal meshes, offer ease of editing but may not be as suitable for rendering high-quality novel views. To harness the strengths of both representations, we propose a new approach that employs a mesh as a guiding mechanism in editing the neural radiance field. We first introduce a differentiable method using marching tetrahedra for polygonal mesh extraction from the neural implicit field and then design a differentiable color extractor to assign colors obtained from the volume renderings to this extracted mesh. This differentiable colored mesh allows gradient back-propagation from the explicit mesh to the implicit fields, empowering users to easily manipulate the geometry and color of neural implicit fields. To enhance user control from coarse-grained to fine-grained levels, we introduce an octree-based structure into its optimization. This structure prioritizes the edited regions and the surface part, making our method achieve fine-grained edits to the neural implicit field and accommodate various user modifications, including object additions, component removals, specific area deformations, and adjustments to local and global colors. Through extensive experiments involving diverse scenes and editing operations, we have demonstrated the capabilities and effectiveness of our method. Our project page is: \url{https://cassiepython.github.io/MNeuEdit/}

cs.CV↗
arXiv2023

Dynamic Dense Graph Convolutional Network for Skeleton-based Human Motion Prediction

Graph Convolutional Networks (GCN) which typically follows a neural message passing framework to model dependencies among skeletal joints has achieved high success in skeleton-based human motion prediction task. Nevertheless, how to construct a graph from a skeleton sequence and how to perform message passing on the graph are still open problems, which severely affect the performance of GCN. To solve both problems, this paper presents a Dynamic Dense Graph Convolutional Network (DD-GCN), which constructs a dense graph and implements an integrated dynamic message passing. More specifically, we construct a dense graph with 4D adjacency modeling as a comprehensive representation of motion sequence at different levels of abstraction. Based on the dense graph, we propose a dynamic message passing framework that learns dynamically from data to generate distinctive messages reflecting sample-specific relevance among nodes in the graph. Extensive experiments on benchmark Human 3.6M and CMU Mocap datasets verify the effectiveness of our DD-GCN which obviously outperforms state-of-the-art GCN-based methods, especially when using long-term and our proposed extremely long-term protocol.

cs.CV↗
arXiv2023

Metal-to-insulator transition in oxide semimetals by anion doping

Oxide semimetals exhibiting both nontrivial topological characteristics stand as exemplary parent compounds and multiple degrees of freedom, offering great promise for the realization of novel electronic states. In this study, we present compelling evidence of profound structural and transport phase shifts in a recently uncovered oxide semimetal, SrNbO3, achieved through effective in-situ anion doping. Notably, a remarkable increase in resistivity of more than three orders of magnitude at room temperature is observed upon nitrogen-doping. The extent of electronic modulation in SrNbO3 is strongly correlated with the misfit strain, underscoring its phase instability to both chemical doping and crystallographic symmetry variations. Using first-principles calculations, we discern that elevating the level of nitrogen doping induces an upward shift in the conductive bands of SrNbO3-dNd. Consequently, a transition from a metallic state to an insulating state becomes apparent as the nitrogen concentration reaches a threshold of 1/3. This investigation sheds light on the potential of anion engineering in oxide semimetals, offering pathways for manipulating their physical properties. These insights hold promise for future applications that harness these materials for tailored functionalities.

cond-mat.mtrl-sci↗
arXiv2023

Strain mediated phase crossover in Ruddlesden Popper nickelates

Recent progress on the signatures of pressure-induced high temperature superconductivity in Ruddlesden Popper (RP) nickelates (Lan+1NinO3n+1) has attracted growing interest in both theoretical calculations and experimental efforts. The fabrication of high-quality single crystalline RP nickelate thin films is critical for possible reducing the superconducting transition pressure and advancing applications in microelectronics in the future. In this study, we report the observations of an active phase transition in RP nickelate films induced by misfit strain. We found that RP nickelate films favor the perovskite structure (n = infinite) under tensile strains, while compressive strains stabilize the La3Ni2O7 (n = 2) phase. The selection of distinct phases is governed by the strain dependent formation energy and electronic configuration. In compressively strained La3Ni2O7, we experimentally determined splitting energy is ~0.2 eV and electrons prefer to occupy in-plane orbitals. First principles calculations unveil a robust coupling between strain effects and the valence state of Ni ions in RP nickelates, suggesting a dual driving force for the inevitable phase co-existence transition in RP nickelates. Our work underscores the sensitivity of RP nickelate formation to epitaxial strain, presenting a significant challenge in fabricating pure-phase RP nickelate films. Therefore, special attention to stacking defects and grain boundaries between different RP phases is essential when discussing the pressure-induced superconductivity in RP nickelates.

cond-mat.supr-con↗
arXiv2023

Regress Before Construct: Regress Autoencoder for Point Cloud Self-supervised Learning

Masked Autoencoders (MAE) have demonstrated promising performance in self-supervised learning for both 2D and 3D computer vision. Nevertheless, existing MAE-based methods still have certain drawbacks. Firstly, the functional decoupling between the encoder and decoder is incomplete, which limits the encoder's representation learning ability. Secondly, downstream tasks solely utilize the encoder, failing to fully leverage the knowledge acquired through the encoder-decoder architecture in the pre-text task. In this paper, we propose Point Regress AutoEncoder (Point-RAE), a new scheme for regressive autoencoders for point cloud self-supervised learning. The proposed method decouples functions between the decoder and the encoder by introducing a mask regressor, which predicts the masked patch representation from the visible patch representation encoded by the encoder and the decoder reconstructs the target from the predicted masked patch representation. By doing so, we minimize the impact of decoder updates on the representation space of the encoder. Moreover, we introduce an alignment constraint to ensure that the representations for masked patches, predicted from the encoded representations of visible patches, are aligned with the masked patch presentations computed from the encoder. To make full use of the knowledge learned in the pre-training stage, we design a new finetune mode for the proposed Point-RAE. Extensive experiments demonstrate that our approach is efficient during pre-training and generalizes well on various downstream tasks. Specifically, our pre-trained models achieve a high accuracy of \textbf{90.28\%} on the ScanObjectNN hardest split and \textbf{94.1\%} accuracy on ModelNet40, surpassing all the other self-supervised learning methods. Our code and pretrained model are public available at: \url{https://github.com/liuyyy111/Point-RAE}.

cs.CV↗
arXiv2023

Non-Hermitian Haldane-Hubbard model: Effective description of one- and two-body dissipation

Using numerically exact diagonalization, we study the correlated Haldane-Hubbard model in the presence of dissipation. Such dissipation can be modeled at short times by the dynamics governed by an effective non-Hermitian Hamiltonian, of which we present a full characterization. If the dissipation corresponds to a two-body loss, the repulsive interaction of the effective Hamiltonian acquires an imaginary component. A competition between the formation of a charge-ordered Mott insulator state and a topological insulator ensues, but with the non-Hermitian contribution aiding in stabilizing the topologically non-trivial regime, delaying the onset of the formation of a local order parameter. Lastly, we analyze the robustness of the ordered phase by following the full dissipative many-body real-time dynamics. An exponentially fast melting of the charge order occurs, whose characteristic rate is roughly independent of the interaction strength, for the case of one-body dissipation.

cond-mat.str-el↗
arXiv2023

Radiative Magnetohydrodynamic Simulation of the Confined Eruption of a Magnetic Flux Rope: Unveiling the Driving and Constraining Forces

We analyse the forces that control the dynamic evolution of a flux rope eruption in a three-dimensional (3D) radiative magnetohydrodynamic (RMHD) simulation. The confined eruption of the flux rope gives rise to a C8.5 flare. The flux rope rises slowly with an almost constant velocity of a few km/s in the early stage, when the gravity and Lorentz force are nearly counterbalanced. After the flux rope rises to the height at which the decay index of the external poloidal field satisfies the torus instability criterion, the significantly enhanced Lorentz force breaks the force balance and drives rapid acceleration of the flux rope. Fast magnetic reconnection is immediately induced within the current sheet under the erupting flux rope, which provides a strong positive feedback to the eruption. The eruption is eventually confined due to the tension force from the strong external toroidal field. Our results suggest that the gravity of plasma plays an important role in sustaining the quasi-static evolution of the pre-eruptive flux rope. The Lorentz force, which is contributed from both the ideal magnetohydrodynamic (MHD) instability and magnetic reconnection, dominates the dynamic evolution during the eruption process.

astro-ph.SR↗
arXiv2023

AvatarCraft: Transforming Text into Neural Human Avatars with Parameterized Shape and Pose Control

Neural implicit fields are powerful for representing 3D scenes and generating high-quality novel views, but it remains challenging to use such implicit representations for creating a 3D human avatar with a specific identity and artistic style that can be easily animated. Our proposed method, AvatarCraft, addresses this challenge by using diffusion models to guide the learning of geometry and texture for a neural avatar based on a single text prompt. We carefully design the optimization framework of neural implicit fields, including a coarse-to-fine multi-bounding box training strategy, shape regularization, and diffusion-based constraints, to produce high-quality geometry and texture. Additionally, we make the human avatar animatable by deforming the neural implicit field with an explicit warping field that maps the target human mesh to a template human mesh, both represented using parametric human models. This simplifies animation and reshaping of the generated avatar by controlling pose and shape parameters. Extensive experiments on various text descriptions show that AvatarCraft is effective and robust in creating human avatars and rendering novel views, poses, and shapes. Our project page is: https://avatar-craft.github.io/.

cs.CV↗
arXiv2023

CDR: Conservative Doubly Robust Learning for Debiased Recommendation

In recommendation systems (RS), user behavior data is observational rather than experimental, resulting in widespread bias in the data. Consequently, tackling bias has emerged as a major challenge in the field of recommendation systems. Recently, Doubly Robust Learning (DR) has gained significant attention due to its remarkable performance and robust properties. However, our experimental findings indicate that existing DR methods are severely impacted by the presence of so-called Poisonous Imputation, where the imputation significantly deviates from the truth and becomes counterproductive. To address this issue, this work proposes Conservative Doubly Robust strategy (CDR) which filters imputations by scrutinizing their mean and variance. Theoretical analyses show that CDR offers reduced variance and improved tail bounds.In addition, our experimental investigations illustrate that CDR significantly enhances performance and can indeed reduce the frequency of poisonous imputation.

cs.IR↗
arXiv2023

Customizing Synthetic Data for Data-Free Student Learning

Data-free knowledge distillation (DFKD) aims to obtain a lightweight student model without original training data. Existing works generally synthesize data from the pre-trained teacher model to replace the original training data for student learning. To more effectively train the student model, the synthetic data shall be customized to the current student learning ability. However, this is ignored in the existing DFKD methods and thus negatively affects the student training. To address this issue, we propose Customizing Synthetic Data for Data-Free Student Learning (CSD) in this paper, which achieves adaptive data synthesis using a self-supervised augmented auxiliary task to estimate the student learning ability. Specifically, data synthesis is dynamically adjusted to enlarge the cross entropy between the labels and the predictions from the self-supervised augmented task, thus generating hard samples for the student model. The experiments on various datasets and teacher-student models show the effectiveness of our proposed method. Code is available at: $\href{https://github.com/luoshiya/CSD}{https://github.com/luoshiya/CSD}$

cs.CV↗
arXiv2023

Adaptive Multi-Teacher Knowledge Distillation with Meta-Learning

Multi-Teacher knowledge distillation provides students with additional supervision from multiple pre-trained teachers with diverse information sources. Most existing methods explore different weighting strategies to obtain a powerful ensemble teacher, while ignoring the student with poor learning ability may not benefit from such specialized integrated knowledge. To address this problem, we propose Adaptive Multi-teacher Knowledge Distillation with Meta-Learning (MMKD) to supervise student with appropriate knowledge from a tailored ensemble teacher. With the help of a meta-weight network, the diverse yet compatible teacher knowledge in the output layer and intermediate layers is jointly leveraged to enhance the student performance. Extensive experiments on multiple benchmark datasets validate the effectiveness and flexibility of our methods. Code is available: https://github.com/Rorozhl/MMKD.

cs.CV↗
arXiv2023

Knowledge Distillation with Deep Supervision

Knowledge distillation aims to enhance the performance of a lightweight student model by exploiting the knowledge from a pre-trained cumbersome teacher model. However, in the traditional knowledge distillation, teacher predictions are only used to provide the supervisory signal for the last layer of the student model, which may result in those shallow student layers lacking accurate training guidance in the layer-by-layer back propagation and thus hinders effective knowledge transfer. To address this issue, we propose Deeply-Supervised Knowledge Distillation (DSKD), which fully utilizes class predictions and feature maps of the teacher model to supervise the training of shallow student layers. A loss-based weight allocation strategy is developed in DSKD to adaptively balance the learning process of each shallow layer, so as to further improve the student performance. Extensive experiments on CIFAR-100 and TinyImageNet with various teacher-student models show significantly performance, confirming the effectiveness of our proposed method. Code is available at: $\href{https://github.com/luoshiya/DSKD}{https://github.com/luoshiya/DSKD}$

cs.LG↗
arXiv2023

Asymmetric Chiral Coupling in a Topological Resonator

Chiral light-matter interactions supported by topological edge modes at the interface of valley photonic crystals provide a robust method to implement the unidirectional spin transfer. The valley topological photonic crystals possess a pair of counterpropagating edge modes. The edge modes are robust against the sharp bend of $60^{\circ}$ and $120^{\circ}$, which can form a resonator with whispering gallery modes. Here, we demonstrate the asymmetric emission of chiral coupling from single quantum dots in a topological resonator by tuning the coupling between a quantum emitter and a resonator mode. Under a magnetic field in Faraday configuration, the exciton state from a single quantum dot splits into two exciton spin states with opposite circularly polarized emissions due to Zeeman effect. Two branches of the quantum dot emissions couple to a resonator mode in different degrees, resulting in an asymmetric chiral emission. Without the demanding of site-control of quantum emitters for chiral quantum optics, an extra degree of freedom to tune the chiral contrast with a topological resonator could be useful for the development of on-chip integrated photonic circuits.

cond-mat.mes-hall↗
arXiv2023

Transport of topologically protected photonic waveguide on chip

We propose a new design on integrated optical devices on-chip with an extra width degree of freedom by using a photonic crystal waveguide with Dirac points between two photonic crystals with opposite valley Chern numbers. With such an extra waveguide, we demonstrate numerically that the topologically protected photonic waveguide keeps properties of valley-locking and immunity to defects. Due to the design flexibility of the width-tunable topologically protected photonic waveguide, many unique on-chip integrated devices have been proposed, such as energy concentrators with a concentration efficiency improvement by more than one order of magnitude, topological photonic power splitter with arbitrary power splitting ratio. The topologically protected photonic waveguide with the width degree of freedom could be beneficial for scaling up photonic devices, which provides a new flexible platform to implement integrated photonic networks on chip.

physics.optics↗
arXiv2023

Syntropic spin alignment at the interface between ferromagnetic and superconducting nitrides

The magnetic correlations at the superconductor/ferromagnet (S/F) interfaces play a crucial role in realizing dissipation-less spin-based logic and memory technologies, such as triplet-supercurrent spin-valves and "π" Josephson junctions. Here we report the coexistence of an induced large magnetic moment and a crypto ferromagnetic state at high-quality nitride S/F interfaces. Using polarized neutron reflectometry and d. c. SQUID measurements, we quantitatively determined the magnetization profile of S/F bilayer and confirmed the induced magnetic moment in the adjacent superconductor only exists below TC. Interestingly, the direction of the induced moment in the superconductors was unexpectedly parallel to that in the ferromagnet, which contrasts with earlier findings in S/F heterostructures based on metals or oxides. The first-principles calculations verify the observed unusual interfacial spin texture is caused by the Heisenberg direct exchange coupling through d orbital overlapping and severe charge transfer across the interfaces. Our work establishes an incisive experimental probe for understanding the magnetic proximity behavior at S/F interfaces and provides a prototype epitaxial building block for superconducting spintronics.

cond-mat.supr-con↗