Article
Computer Science, Theory & Methods
Bao Gia Doan, Minhui Xue, Shiqing Ma, Ehsan Abbasnejad, Damith C. Ranasinghe
Summary: Deep neural networks are vulnerable to attacks from adversarial inputs and Trojans. This study introduces a class of spatially bounded, physically realizable adversarial examples called Universal NaTuralistic adversarial paTches (TnTs). TnTs are highly effective and universal, allowing an attacker to exert a greater level of control and deploy patches in the physical world. Extensive experiments demonstrate the realistic threat from TnTs and their robustness against state-of-the-art deep neural networks.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2022)
Article
Computer Science, Artificial Intelligence
Bulent Nafi Ornek, Salih Berkan Aydemir, Timur Duzenli, Bilal Ozak
Summary: This study proposes a complex valued activation function for classification problems with complex valued data, and analyzes its performance using a complex-valued extreme learning classifier. The proposed activation function is obtained by performing extremal analyses of the inequalities derived from the boundary Schwarz lemma. Simulation results show that the proposed activation function outperforms other activation functions in terms of classification accuracy and function approximation.
Article
Computer Science, Information Systems
Seung-Yeon Hwang, Jeong-Joon Kim
Summary: Recently, deep learning has made significant progress in areas that require human cognitive ability, learning ability, and reasoning ability. Activation functions play a crucial role in enabling artificial neural networks to learn complex patterns through nonlinearity. Different activation functions are being researched to address issues such as vanishing gradients and dying nodes that may arise during deep learning. This paper proposes a universal activation function (UA) that allows researchers to easily create and apply various activation functions and enhance the performance of neural networks. Through proper adjustment of three hyperparameters, UA can generate new types of activation functions as well as functions similar to traditional ones. The experimental evaluation using Convolutional Neural Network (CNN) and benchmark dataset shows that UA improves the classification performance of CNNs by up to 5% compared to traditional activation functions in most cases.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Physics, Multidisciplinary
S. M. Sivalingam, Pushpendra Kumar, V. Govindaraj
Summary: In this paper, a neural network-based approach with an Extreme Learning Machine (ELM) is proposed for solving fractional differential equations. The solution procedure for both linear and nonlinear fractional differential equations is derived, and the convergence and stability of the proposed method are investigated. Numerical solutions for several fractional-order ordinary and partial differential equations are examined, and the impact of changing the number of neurons on solution accuracy is graphically determined.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2023)
Article
Automation & Control Systems
Jianwen Guo, Xiaoyan Li, Zhiyuan Liu, Shaohui Zhang, Jiapeng Wu, Chuan Li, Jianyu Long
Summary: The study introduces a new activation function, doublet ELM (DELM), evaluates its performance using experimental data collected from a 3D printer platform, and compares it with other activation functions. Results show that DELM achieves the highest accuracy in different hidden nodes and performs well even with small samples.
Article
Computer Science, Information Systems
Shao-Bo Lin, Kaidong Wang, Yao Wang, Ding-Xuan Zhou
Summary: Compared to practical research activities, the study of theoretical behaviors of deep convolutional neural networks (DCNNs) is significantly lacking behind. However, this paper proves that implementing empirical risk minimization on DCNNs with expansive convolution can be strongly universally consistent. Through a series of experiments, it is shown that DCNNs with expansive convolution, even without fully connected layers, perform as well as deep neural networks with contracting convolutional layers and fully connected layers.
IEEE TRANSACTIONS ON INFORMATION THEORY
(2022)
Article
Computer Science, Artificial Intelligence
Siwar Yahia, Salwa Said, Mourad Zaied
Summary: This paper proposes a new structure based on WNN, deep architecture, and ELM, which improves the classification accuracy in machine learning applications by using a composite wavelet activation function and an ELM auto-encoder with DL structure.
Article
Biochemistry & Molecular Biology
Prakarsh Yadav, Parisa Mollaei, Zhonglin Cao, Yuyang Wang, Amir Barati Farimani
Summary: This study developed three machine learning approaches to predict the conformation state and activity level of GPCR proteins. These approaches accurately predict the activation state of GPCRs and help determine important features.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Review
Computer Science, Artificial Intelligence
Andrea Apicella, Francesco Donnarumma, Francesco Isgro, Roberto Prevete
Summary: In recent years, there has been a renewed interest in trainable activation functions, which can be trained during the learning process to improve neural network performance. Various models of trainable activation functions have been proposed in the literature, many of which are equivalent to adding neuron layers with fixed activation functions and simple local rules.
Article
Computer Science, Information Systems
Massimo Vatalaro, Tatiana Moposita, Sebastiano Strangio, Lionel Trojman, Andrei Vladimirescu, Marco Lanuzza, Felice Crupi
Summary: This paper presents a novel low-power low-voltage analog implementation of the softmax function, with electrically adjustable amplitude and slope parameters. The proposed modular design allows scalability based on the number of inputs and outputs, and the circuit, designed in a 0.18μm CMOS technology, operates efficiently at a supply voltage of 500mV. The compact and energy-efficient option consumes only 3μW of power for a ten-input/ten-output realization, showcasing its potential in comparison to digital implementations.
Article
Multidisciplinary Sciences
Mitsumasa Nakajima, Katsuma Inoue, Kenji Tanaka, Yasuo Kuniyoshi, Toshikazu Hashimoto, Kohei Nakajima
Summary: The research presents a physical deep learning approach that can train physical neural networks without knowledge of the physical system and its gradient. Through the use of an optoelectronic recurrent neural network, the concept was validated with competitive accelerated computation performance on benchmarks.
NATURE COMMUNICATIONS
(2022)
Article
Mathematics, Applied
Naxian Ni, Suchuan Dong
Summary: This paper introduces a modified ELM method, called HLConcELM, which can produce highly accurate solutions to linear/nonlinear PDEs when the last hidden layer of the network is narrow and when it is wide. The method is based on a modified feedforward neural network (FNN), termed HLConcFNN, which incorporates a logical concatenation of the hidden layers in the network and exposes all the hidden nodes to the output-layer nodes. The HLConcFNNs have the interesting property that the representation capacity of the network is guaranteed to be not smaller than that of the original network architecture, even with additional hidden layers or nodes.
JOURNAL OF SCIENTIFIC COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Krishnan Raghavan, Shweta Garg, Sarangapani Jagannathan, V. A. Samaranayake
Summary: This article introduces a novel learning methodology for classification in high-dimensional data, addressing challenges through a L-1 regularized zero-sum game. The proposed approach utilizes distributed learning and an alternating minimization method for optimal sparsity. The efficiency of the approach is demonstrated theoretically and empirically with nine data sets.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Article
Thermodynamics
Matthew T. Hughes, Brian M. Fronk, Srinivas Garimella
Summary: A novel universal model has been proposed to predict the condensation frictional pressure drop and heat transfer coefficient for horizontal microchannel and macrochannel flows. Machine learning algorithms, especially the random forest model, were found to predict the data best for pressure drop and heat transfer.
INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER
(2021)
Article
Automation & Control Systems
Mohammad Rezaei-Ravari, Mahdi Eftekhari, Farid Saberi-Movahed
Summary: Multi-label learning methods are regularized via Locally Linear Embedding (LLE) to increase efficiency, with experiments showing that using dual-manifold learning as the training method for neural classifiers significantly improves classification performance.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Zhihai Yang, Lin Xu, Zhongmin Cai, Zongben Xu
KNOWLEDGE-BASED SYSTEMS
(2016)
Article
Computer Science, Artificial Intelligence
Lin Xu, Shaobo Lin, Yao Wang, Zongben Xu
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2017)
Article
Automation & Control Systems
Lin Xu, Shaobo Lin, Jinshan Zeng, Xia Liu, Yi Fang, Zongben Xu
IEEE TRANSACTIONS ON CYBERNETICS
(2018)
Article
Geochemistry & Geophysics
Jing Yao, Danfeng Hong, Lin Xu, Deyu Meng, Jocelyn Chanussot, Zongben Xu
Summary: This article presents a novel blind hyperspectral unmixing (HU) model called sparsity-enhanced convolutional decomposition (SeCoDe), which captures spatial-spectral information of hyperspectral imagery (HSI) in a tensor-based fashion. The model effectively addresses the ill-posed problems and spectral variability in HSI, resulting in superior unmixing performance compared to previous methods.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Shiyang Yan, Jianan Zhao, Lin Xu
Summary: This paper proposes an adaptive multi-task learning scheme for cross domain and modal person re-identification, which effectively utilizes visual and language information from multiple datasets to improve learning performance. The proposed method models the domain difference and the relationship between vision and language modalities, achieving state-of-the-art performance.
Article
Engineering, Multidisciplinary
Lin Xu, Xiangyong Cao, Jing Yao, Zheng Yan
Summary: In this paper, we propose an Orthogonal Super Greedy learning (OSGL) method for hidden neurons selection in feedforward neural networks. The method addresses the issue of generation performance and computational complexity being affected by irrelevant hidden variables. Theoretical analyses and empirical results demonstrate that the proposed method can achieve optimal learning rate and produce excellent generalization performance with a sparse and compact feature representation.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Lifan Zhao, Yunlong Meng, Lin Xu
Summary: In this paper, we propose a framework called OA-FSUI2IT for addressing the few-shot cross domain object detection task with limited unlabeled images in the target domain. The framework includes a discriminator augmentation module, a patch pyramid contrastive learning strategy, and a self-supervised content-consistency loss. Extensive experiments demonstrate the effectiveness of our framework for FSCD object detection, achieving state-of-the-art performance on benchmark tests.
THIRTY-SIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FOURTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE / THE TWELVETH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Jianan Zhao, Fengliang Qi, Guangyu Ren, Lin Xu
Summary: Vehicle re-identification plays a crucial role in urban operation, management, and security, facing the challenges of utilizing raw data and learning robust re-ID models. This paper introduces a video vehicle re-ID benchmark VVeRI-901 and proposes a new PhD learning method to address these challenges, demonstrating the superiority of video-based re-ID over image-based re-ID.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021
(2021)
Article
Computer Science, Artificial Intelligence
Haixia Bi, Lin Xu, Xiangyong Cao, Yong Xue, Zongben Xu
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Han Sun, Zhiyuan Chen, Shiyang Yan, Lin Xu
2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019)
(2019)
Proceedings Paper
Computer Science, Artificial Intelligence
Lin Xu, Han Sun, Yuai Liu
2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019)
(2019)
Article
Computer Science, Artificial Intelligence
Rui Lv, Dingheng Wang, Jiangbin Zheng, Zhao-Xu Yang
Summary: In this paper, the authors investigate tensor decomposition for neural network compression. They analyze the convergence and precision of tensor mapping theory, validate the rationality of tensor mapping and its superiority over traditional tensor approximation based on the Lottery Ticket Hypothesis. They propose an efficient method called 3D-KCPNet to compress 3D convolutional neural networks using the Kronecker canonical polyadic (KCP) tensor decomposition. Experimental results show that 3D-KCPNet achieves higher accuracy compared to the original baseline model and the corresponding tensor approximation model.
Article
Computer Science, Artificial Intelligence
Xiangkun He, Zhongxu Hu, Haohan Yang, Chen Lv
Summary: In this paper, a novel constrained multi-objective reinforcement learning algorithm is proposed for personalized end-to-end robotic control with continuous actions. The approach trains a single model using constraint design and a comprehensive index to achieve optimal policies based on user-specified preferences.
Article
Computer Science, Artificial Intelligence
Zhijian Zhuo, Bilian Chen, Shenbao Yu, Langcai Cao
Summary: In this paper, a novel method called Expansion with Contraction Method for Overlapping Community Detection (ECOCD) is proposed, which utilizes non-negative matrix factorization to obtain disjoint communities and applies expansion and contraction processes to adjust the degree of overlap. ECOCD is applicable to various networks with different properties and achieves high-quality overlapping community detection.
Article
Computer Science, Artificial Intelligence
Yizhe Zhu, Chunhui Zhang, Jialin Gao, Xin Sun, Zihan Rui, Xi Zhou
Summary: In this work, the authors propose a Contrastive Spatio-Temporal Distilling (CSTD) approach to improve the detection of high-compressed deepfake videos. The approach leverages spatial-frequency cues and temporal-contrastive alignment to fully exploit spatiotemporal inconsistency information.
Review
Computer Science, Artificial Intelligence
Laijin Meng, Xinghao Jiang, Tanfeng Sun
Summary: This paper provides a review of coverless steganographic algorithms, including the development process, known contributions, and general issues in image and video algorithms. It also discusses the security of coverless steganography from theoretical analysis to actual investigation for the first time.
Article
Computer Science, Artificial Intelligence
Yajie Bao, Tianwei Xing, Xun Chen
Summary: Visual question answering requires processing multi-modal information and effective reasoning. Neural-symbolic learning is a promising method, but current approaches lack uncertainty handling and can only provide a single answer. To address this, we propose a confidence based neural-symbolic approach that evaluates NN inferences and conducts reasoning based on confidence.
Article
Computer Science, Artificial Intelligence
Anh H. Vo, Bao T. Nguyen
Summary: Interior style classification is an interesting problem with potential applications in both commercial and academic domains. This project proposes a method named ISC-DeIT, which combines data-efficient image transformer architectures and knowledge distillation, to address the interior style classification problem. Experimental results demonstrate a significant improvement in predictive accuracy compared to other state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Shashank Kotyan, Danilo Vasconcellos Vargas
Summary: This article introduces a novel augmentation technique called Dynamic Scanning Augmentation to improve the accuracy and robustness of Vision Transformer (ViT). The technique leverages dynamic input sequences to adaptively focus on different patches, resulting in significant changes in ViT's attention mechanism. Experimental results demonstrate that Dynamic Scanning Augmentation outperforms ViT in terms of both robustness to adversarial attacks and accuracy against natural images.
Article
Computer Science, Artificial Intelligence
Hiba Alqasir, Damien Muselet, Christophe Ducottet
Summary: The article proposes a solution to improve the learning process of a classification network by providing shape priors, reducing the need for annotated data. The solution is tested on cross-domain digit classification tasks and a video surveillance application.
Article
Computer Science, Artificial Intelligence
Dexiu Ma, Mei Liu, Mingsheng Shang
Summary: This paper proposes a method using neural dynamics solvers to solve infinity-norm optimization problems. Two improved solvers are constructed and their effectiveness and superiority are demonstrated through theoretical analysis and simulation experiments.
Article
Computer Science, Artificial Intelligence
Francesco Gregoretti, Giovanni Pezzulo, Domenico Maisto
Summary: Active Inference is a computational framework that uses probabilistic inference and variational free energy minimization to describe perception, planning, and action. cpp-AIF is a header-only C++ library that provides a powerful tool for implementing Active Inference for Partially Observable Markov Decision Processes through multi-core computing. It is cross-platform and improves performance, memory management, and usability compared to existing software.
Article
Computer Science, Artificial Intelligence
Zelin Ying, Dawei Cheng, Cen Chen, Xiang Li, Peng Zhu, Yifeng Luo, Yuqi Liang
Summary: This paper proposes a novel stock market trends prediction framework called SMART, which includes a self-supervised stock technical data sequence embedding model S3E. By training with multiple self-supervised auxiliary tasks, the model encodes stock technical data sequences into embeddings and uses the learned sequence embeddings for predicting stock market trends. Extensive experiments on China A-Shares market and NASDAQ market prove the high effectiveness of our model in stock market trends prediction, and its effectiveness is further validated in real-world applications in a leading financial service provider in China.
Article
Computer Science, Artificial Intelligence
Hao Li, Hao Jiang, Dongsheng Ye, Qiang Wang, Liang Du, Yuanyuan Zeng, Liu Yuan, Yingxue Wang, C. Chen
Summary: DHGAT1, a dynamic hyperbolic graph attention network, utilizes hyperbolic metric properties to embed dynamic graphs. It employs a spatiotemporal self-attention mechanism and weighted node representations, resulting in excellent performance in link prediction tasks.
Article
Computer Science, Artificial Intelligence
Jiehui Huang, Zhenchao Tang, Xuedong He, Jun Zhou, Defeng Zhou, Calvin Yu-Chian Chen
Summary: This study proposes a progressive learning multi-scale feature blending model for image deraining tasks. The model utilizes detail dilation and texture extraction to improve the restoration of rainy images. Experimental results show that the model achieves near state-of-the-art performance in rain removal tasks and exhibits better rain removal realism.
Article
Computer Science, Artificial Intelligence
Lizhi Liu, Zilin Gao, Yinhe Wang, Yongfu Li
Summary: This paper proposes a novel discrete-time interconnected model for depicting complex dynamical networks. The model consists of nodes and edges subsystems, which consider the dynamic characteristic of both nodes and edges. By designing control strategies and coupling modes, the stabilization and synchronization of the network are achieved. Simulation results demonstrate the effectiveness of the proposed methods.