查看更多>>摘要：Current synthetic aperture radar (SAR) ship classifiers using convolutional neural networks (CNNs) offer state-of-the-art performance. Yet, they still have two defects potentially hindering accuracy progress - polarization insufficient utilization and traditional feature abandonment. Therefore, we propose a polar-ization fusion network with geometric feature embedding (PFGFE-Net) to solve them. PFGFE-Net achieves the polarization fusion (PF) from the input data, feature-level, and decision-level. Moreover, the geometric feature embedding (GFE) enriches expert experience. Results on OpenSARShip reveal PFGFE-Net's excel-lent performance. (c) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：Least squares regression (LSR) is an important machine learning method for feature extraction, feature selection, and image classification. For the training samples, there are correlations among samples from the same class. Therefore, many LSR-based methods utilize this property to pursue discriminative representation. However, if the training samples contain noise or outliers, it will be hard to obtain the exact inter-class correlation. To address this problem, in this paper, a novel LSR-based method is proposed, named low-rank inter-class sparsity based semi-flexible target least squares regression (LIS_StLSR). Firstly, the low-rank representation method is utilized to achieve the intrinsic characteristics of the training samples. Afterwards, the low-rank inter-class sparsity constraint is used to force the projected data to have an exact common sparsity structure in each class, which will be robust to noise and outliers in the training samples. This step can also reduce margins of samples from the same class and enlarge margins of samples from different classes to make the projection matrix discriminative. The low-rank representation and the discriminative projection matrix are jointly learned such that they can be boosted mutually. Moreover, a semi-flexible regression target matrix is introduced to measure the regression error more accurately, thus the regression performance can be enhanced to improve the classification accuracy. Experiments are implemented on the different databases of Yale B, AR, LFW, CASIA NIR-VIS, 15-Scene SPF, COIL-20, and Caltech 101, illustrating that the proposed LIS_StLSR outperforms many state-of-the-art methods. (c) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：An Orthogonal Least Squares (OLS) based feature selection method is proposed for both binomial and multinomial classification. The novel Squared Orthogonal Correlation Coefficient (SOCC) is defined based on Error Reduction Ratio (ERR) in OLS and used as the feature ranking criterion. The equivalence between the canonical correlation coefficient, Fisher's criterion, and the sum of the SOCCs is revealed, which unveils the statistical implication of ERR in OLS for the first time. It is also shown that the OLS based feature selection method has speed advantages when applied for greedy search. The proposed method is comprehensively compared with the mutual information based feature selection methods and the embedded methods using both synthetic and real world datasets. The results show that the proposed method is always in the top 5 among the 12 candidate methods. Besides, the proposed method can be directly applied to continuous features without discretisation, which is another significant advantage over mutual information based methods.(c) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：Cutting plane methods play a significant role in modern solvers for tackling mixed-integer programming (MIP) problems. Proper selection of cuts would remove infeasible solutions in the early stage, thus largely reducing the computational burden without hurting the solution accuracy. However, the major cut selection approaches heavily rely on heuristics, which strongly depend on the specific problem at hand and thus limit their generalization capability. In this paper, we propose a data-driven and generalizable cut selection approach, named CUT RANKING , in the settings of multiple instance learning. To measure the quality of the candidate cuts, a scoring function, which takes the instance-specific cut features as inputs, is trained and applied in cut ranking and selection. In order to evaluate our method, we conduct extensive experiments on both synthetic datasets and real-world datasets. Compared with commonly used heuristics for cut selection, the learning-based policy has shown to be more effective, and is capable of generalizing over multiple problems with different properties. CUT RANKING has been deployed in an industrial solver for large-scale MIPs. In the online A/B testing of the product planning problems with more than 10 7 variables and constraints daily, CUT RANKING has achieved the average speedup ratio of 12.42% over the production solver without any accuracy loss of solution. (C) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：Incomplete multi-view clustering partitions multi-view data suffering from missing views, for which ma-trix factorization approaches seek the latent representation of incomplete multi-view data and constitute one effective category of methods. To exploit data properties further, manifold structure preserving is also incorporated into matrix factorization. However, previous methods optimized the data similarity matrix in the manifold structure preserving term as an unknown variable, which is not guaranteed to faithfully represent the similarities of the original multi-view data and also increases the computational difficulty. To overcome these drawbacks, in this paper, we propose Incomplete Multi-view Clustering with Cosine Similarity (IMCCS). In IMCCS, we directly calculate the cosine similarity in the original multi-view space to strengthen the ability of preserving the manifold structure of the original multi-view data. There is no need to introduce the additional variable. The manifold structure preserving term with cosine similarity and the matrix factorization term are integrated into a unified objective function. An iterative algorithm with gradient descent is designed to solve this objective. Extensive experiments on multi-view datasets show that IMCCS outperforms state-of-the-art incomplete multi-view clustering methods. (c) 2021 Elsevier Ltd. All rights reserved.
Obeso, Abraham MontoyaBenois-Pineau, JennyVazquez, Mireya Sarai GarciaAcosta, Alejandro alvaro Ramirez...
14页
查看更多>>摘要：The so-called "attention mechanisms" in Deep Neural Networks (DNNs) denote an automatic adaptation of DNNs to capture representative features given a specific classification task and related data. Such atten-tion mechanisms perform both globally by reinforcing feature channels and locally by stressing features in each feature map. Channel and feature importance are learnt in the global end-to-end DNNs train-ing process. In this paper, we present a study and propose a method with a different approach, adding supplementary visual data next to training images. We use human visual attention maps obtained inde-pendently with psycho-visual experiments, both in task-driven or in free viewing conditions, or powerful models for prediction of visual attention maps. We add visual attention maps as new data alongside im-ages, thus introducing human visual attention into the DNNs training and compare it with both global and local automatic attention mechanisms. Experimental results show that known attention mechanisms in DNNs work pretty much as human visual attention, but still the proposed approach allows a faster convergence and better performance in image classification tasks.(c) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：The field of electrocardiogram (ECG) biometrics has received considerable attention in recent years. Although some promising methods have been proposed, it is challenging to design a robust and precise method to improve the recognition performance of ECG signals with noise and sample variation. While the advantage of improved local binary pattern (LBP) for establishing identities has been widely recognized, extracting the latent semantics from multiple LBP features has attracted little attention. We propose a robust multi-feature collective non-negative matrix factorization (RMCNMF) model to handle noise and sample variation in ECG Biometrics. We extract multiple LBP histograms as feature descriptors from segmented ECG signals, and propose a multi-feature learning framework that learns unified representations in the shared latent semantic space via collective non-negative matrix factorization. To further enhance the discrimination of learned representations, we integrate label information and multiple norms in the proposed model, which not only preserves intra-and inter-subject similarities but also mitigates the influence of noise and sample variation. RMCNMF can be solved by an efficient iteration method, for which we provide a convergence analysis in detail. Extensive experiments on four ECG databases show that it performs competitively with state-of-the-art methods. (c) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：Dynamic-to-static image translation aims to convert the dynamic scene into static so that dynamic elements are eliminated from the image. Recent works typically see the problem as an image-to-image translation task, and perform the learned feature mapping over the whole dynamic image to synthesize the static image, which leads to unnecessary detail loss in original static regions. To that end, we delicately formulate it as an image inpainting-like problem to fill the missing static pixels in dynamic regions while retaining original static regions. We achieve this by proposing a coarse-to-fine framework. At coarse stage, we utilize a simple encoder-decoder network to rough out the static image. Using the coarse predicted image, we explicitly infer a more accurate dynamic mask to identify both dynamic objects and their shadows, so that the task could be effectively converted to an image inpainting problem. At fine stage, we recover the missing static pixels in the estimated dynamic regions on the basis of their coarse predictions. We enhance the coarse predicted contents by proposing a mutual texture-structure attention module, which enables the dynamic regions to borrow textures and structures separately from distant locations based on contextual similarity. Several losses are combined as the training objective function to generate excellent results with global consistency and fine details. Qualitative and quantitative experiments verify the superiority of our method in restoring high-quality static contents over state-of-the-art models. In addition, we evaluate the usefulness of the recovered static images by using them as query images to improve visual place recognition in dynamic scenes. (c) 2021 Elsevier Ltd. All rights reserved.
查看更多>>摘要：In reinforcement learning (RL), the intrinsic reward estimation is necessary for policy learning when the extrinsic reward is sparse or absent. To this end, Unified Curiosity-driven Learning with Smoothed intrinsic reward Estimation (UCLSE) is proposed to address the sparse extrinsic reward problem from the perspective of completeness of intrinsic reward estimation. We further propose state distribution-aware weighting method and policy-aware weighting method to dynamically unify two mainstream intrinsic reward estimation methods. In this way, the agent can explore the environment more effectively and efficiently. Under this framework, we propose to employ an attention module to extract task-relevant features for a more precise estimation of intrinsic reward. Moreover, we propose to improve the robustness of policy learning by smoothing the intrinsic reward with a batch of transitions close to the current transition. Extensive experimental results on Atari games demonstrate that our method outperforms the state-of-the-art approaches in terms of both score and training efficiency. (c) 2021 Published by Elsevier Ltd.
查看更多>>摘要：Surface defect segmentation is very important for the quality inspection of industrial production and is an important pattern recognition problem. Although deep learning (DL) has achieved remarkable results in surface defect segmentation, most of these results have been obtained by using massive images with pixel-level annotations, which are difficult to obtain at industrial sites. This paper proposes a weakly supervised defect segmentation method based on the dynamic templates generated by an improved cycle consistent generative adversarial network (CycleGAN) trained by image-level annotations. To generate better templates for defects with weak signals, we propose a defect attention module by applying the defect residual for the discriminator to strengthen the elimination of defect regions and suppress changes in the background. A defect cycle-consistent loss is designed by adding structural similarity (SSIM) to the original L1 loss to include the grayscale and structural features; the proposed loss can better model the inner structure of defects. After obtaining the defect-free template, a defect segmentation map can easily be obtained through a simple image comparison and threshold segmentation. Experiments show that the proposed method is both efficient and effective, significantly outperforms other weakly supervised methods, and achieves performance that is comparable or even superior to that of supervised methods on three industrial datasets (intersection over union (IoU) on the DAGM 2007, KSD and CCSD datasets of 78.28%, 59.43%,and 68.83%, respectively). The proposed method can also be employed as a semiautomatic annotation tool combined with active learning.(c) 2021 Elsevier Ltd. All rights reserved.
NSTL
Elsevier