Search Results for author: Xian Wei

Found 32 papers, 4 papers with code

WaveAttack: Asymmetric Frequency Obfuscation-based Backdoor Attacks Against Deep Neural Networks

no code implementations17 Oct 2023 Jun Xia, Zhihao Yue, Yingbo Zhou, Zhiwei Ling, Xian Wei, Mingsong Chen

Due to the popularity of Artificial Intelligence (AI) technology, numerous backdoor attacks are designed by adversaries to mislead deep neural network predictions by manipulating training samples and training processes.

Backdoor Attack SSIM

Continual Learning via Manifold Expansion Replay

no code implementations12 Oct 2023 Zihao Xu, Xuan Tang, Yufei Shi, Jianfeng Zhang, Jian Yang, Mingsong Chen, Xian Wei

To address this problem, we propose a novel replay strategy called Manifold Expansion Replay (MaER).

Continual Learning Management

EqGAN: Feature Equalization Fusion for Few-shot Image Generation

no code implementations27 Jul 2023 Yingbo Zhou, Zhihao Yue, Yutong Ye, Pengyu Zhang, Xian Wei, Mingsong Chen

Due to the absence of fine structure and texture information, existing fusion-based few-shot image generation methods suffer from unsatisfactory generation quality and diversity.

Generative Adversarial Network Image Generation

CTAGE: Curvature-Based Topology-Aware Graph Embedding for Learning Molecular Representations

no code implementations25 Jul 2023 Yili Chen, Zhengyu Li, Zheng Wan, Hui Yu, Xian Wei

Therefore, it is necessary to develop a method for predicting molecular properties that effectively combines spatial structural information while maintaining the simplicity and efficiency of graph neural networks.

Graph Embedding Molecular Property Prediction +1

Hyperbolic Graph Diffusion Model

1 code implementation13 Jun 2023 Lingfeng Wen, Xuan Tang, Mingjie Ouyang, Xiangxiang Shen, Jian Yang, Daxin Zhu, Mingsong Chen, Xian Wei

In order to simultaneously utilize the data generation capabilities of diffusion models and the ability of hyperbolic embeddings to extract latent hierarchical distributions, we propose a novel graph generation method called, Hyperbolic Graph Diffusion Model (HGDM), which consists of an auto-encoder to encode nodes into successive hyperbolic embeddings, and a DM that operates in the hyperbolic latent space.

Graph Generation

Simplicial Message Passing for Chemical Property Prediction

no code implementations9 Jun 2023 Hai Lan, Xian Wei

Recently, message-passing Neural networks (MPNN) provide a promising tool for dealing with molecular graphs and have achieved remarkable success in facilitating the discovery and materials design with desired properties.

Property Prediction

FedMR: Federated Learning via Model Recombination

no code implementations18 May 2023 Ming Hu, Zhihao Yue, Zhiwei Ling, Yihao Huang, Cheng Chen, Xian Wei, Yang Liu, Mingsong Chen

Although Federated Learning (FL) enables global model training across clients without compromising their raw data, existing Federated Averaging (FedAvg)-based methods suffer from the problem of low inference performance, especially for unevenly distributed data among clients.

Federated Learning

Group Equivariant BEV for 3D Object Detection

no code implementations26 Apr 2023 Hongwei Liu, Jian Yang, Jianfeng Zhang, Dongheng Shao, Jielong Guo, Shaobo Li, Xuan Tang, Xian Wei

Experimental results demonstrate that GeqBevNet can extract more rotational equivariant features in the 3D object detection of the actual road scene and improve the performance of object orientation prediction.

3D Object Detection Object +2

Autoencoders with Intrinsic Dimension Constraints for Learning Low Dimensional Image Representations

no code implementations16 Apr 2023 Jianzhang Zheng, Hao Shen, Jian Yang, Xuan Tang, Mingsong Chen, Hui Yu, Jielong Guo, Xian Wei

Motivated by the important role of ID, in this paper, we propose a novel deep representation learning approach with autoencoder, which incorporates regularization of the global and local ID constraints into the reconstruction of data representations.

Image Classification Representation Learning

A Survey of Geometric Optimization for Deep Learning: From Euclidean Space to Riemannian Manifold

no code implementations16 Feb 2023 Yanhong Fei, Xian Wei, Yingjie Liu, Zhengyu Li, Mingsong Chen

Although Deep Learning (DL) has achieved success in complex Artificial Intelligence (AI) tasks, it suffers from various notorious problems (e. g., feature redundancy, and vanishing or exploding gradients), since updating parameters in Euclidean space cannot fully exploit the geometric structure of the solution space.

Transfer Learning

CyclicFL: A Cyclic Model Pre-Training Approach to Efficient Federated Learning

no code implementations28 Jan 2023 Pengyu Zhang, Yingbo Zhou, Ming Hu, Xin Fu, Xian Wei, Mingsong Chen

Based on the concept of Continual Learning (CL), we prove that CyclicFL approximates existing centralized pre-training methods in terms of classification and prediction performance.

Continual Learning Federated Learning

HierarchyFL: Heterogeneous Federated Learning via Hierarchical Self-Distillation

no code implementations5 Dec 2022 Jun Xia, Yi Zhang, Zhihao Yue, Ming Hu, Xian Wei, Mingsong Chen

Federated learning (FL) has been recognized as a privacy-preserving distributed machine learning paradigm that enables knowledge sharing among various heterogeneous artificial intelligence (AIoT) devices through centralized global model aggregation.

Federated Learning Privacy Preserving

Continual Learning for Pose-Agnostic Object Recognition in 3D Point Clouds

no code implementations11 Sep 2022 Xihao Wang, Xian Wei

Continual Learning aims to learn multiple incoming new tasks continually, and to keep the performance of learned tasks at a consistent level.

Continual Learning Object Recognition

FedMR: Fedreated Learning via Model Recombination

no code implementations16 Aug 2022 Ming Hu, Zhihao Yue, Zhiwei Ling, Xian Wei, Mingsong Chen

Worse still, in each round of FL training, FedAvg dispatches the same initial local models to clients, which can easily result in stuck-at-local-search for optimal global models.

Federated Learning Privacy Preserving

Model-Contrastive Learning for Backdoor Defense

1 code implementation9 May 2022 Zhihao Yue, Jun Xia, Zhiwei Ling, Ming Hu, Ting Wang, Xian Wei, Mingsong Chen

Due to the popularity of Artificial Intelligence (AI) techniques, we are witnessing an increasing number of backdoor injection attacks that are designed to maliciously threaten Deep Neural Networks (DNNs) causing misclassification.

Backdoor Attack backdoor defense +1

Eliminating Backdoor Triggers for Deep Neural Networks Using Attention Relation Graph Distillation

1 code implementation21 Apr 2022 Jun Xia, Ting Wang, Jiepin Ding, Xian Wei, Mingsong Chen

Due to the prosperity of Artificial Intelligence (AI) techniques, more and more backdoors are designed by adversaries to attack Deep Neural Networks (DNNs). Although the state-of-the-art method Neural Attention Distillation (NAD) can effectively erase backdoor triggers from DNNs, it still suffers from non-negligible Attack Success Rate (ASR) together with lowered classification ACCuracy (ACC), since NAD focuses on backdoor defense using attention features (i. e., attention maps) of the same order.

backdoor defense Knowledge Distillation +1

Learning from Attacks: Attacking Variational Autoencoder for Improving Image Classification

no code implementations11 Mar 2022 Jianzhang Zheng, Fan Yang, Hao Shen, Xuan Tang, Mingsong Chen, Liang Song, Xian Wei

We propose an algorithmic framework that leverages the advantages of the DNNs for data self-expression and task-specific predictions, to improve image classification.

Classification Image Classification

O-ViT: Orthogonal Vision Transformer

no code implementations28 Jan 2022 Yanhong Fei, Yingjie Liu, Xian Wei, Mingsong Chen

Inspired by the tremendous success of the self-attention mechanism in natural language processing, the Vision Transformer (ViT) creatively applies it to image patch sequences and achieves incredible performance.

Synthesizing Tensor Transformations for Visual Self-attention

no code implementations5 Jan 2022 Xian Wei, Xihao Wang, Hai Lan, JiaMing Lei, Yanhui Huang, Hui Yu, Jian Yang

Self-attention shows outstanding competence in capturing long-range relationships while enhancing performance on vision tasks, such as image classification and image captioning.

Image Captioning Image Classification

ViR:the Vision Reservoir

no code implementations27 Dec 2021 Xian Wei, Bin Wang, Mingsong Chen, Ji Yuan, Hai Lan, Jiehuang Shi, Xuan Tang, Bo Jin, Guozhang Chen, Dongping Yang

To address these problems, a novel method, namely, Vision Reservoir computing (ViR), is proposed here for image classification, as a parallel to ViT.

Classification Image Classification

Couplformer:Rethinking Vision Transformer with Coupling Attention Map

1 code implementation10 Dec 2021 Hai Lan, Xihao Wang, Xian Wei

With the development of the self-attention mechanism, the Transformer model has demonstrated its outstanding performance in the computer vision domain.

Image Classification

Boost Neural Networks by Checkpoints

no code implementations NeurIPS 2021 Feng Wang, Guoyizhe Wei, Qiao Liu, Jinxiang Ou, Xian Wei, Hairong Lv

In the experiments, it yields up to 5. 02% higher accuracy over single EfficientNet-B0 on the imbalanced datasets.

Trace Quotient with Sparsity Priors for Learning Low Dimensional Image Representations

no code implementations8 Oct 2018 Xian Wei, Hao Shen, Martin Kleinsteuber

We propose a generic algorithmic framework, which leverages two classic representation learning paradigms, i. e., sparse representation and the trace quotient criterion.

Data Visualization Dimensionality Reduction +1

An Adaptive Dictionary Learning Approach for Modeling Dynamical Textures

no code implementations19 Dec 2013 Xian Wei, Hao Shen, Martin Kleinsteuber

Video representation is an important and challenging task in the computer vision community.

Dictionary Learning

Cannot find the paper you are looking for? You can Submit a new open access paper.