Search

• Almost all these works above take the independent and identically distributed (i.i.d.) assumption, i.e., regarding all samples in various computer vision tasks as i.i.d., but they are not i.i.d. samples and the samples of many machine learning applications such as speech recognition, system diagnosis, and market forecasting are also proved to be non-i.i.d. samples. In particular, with the advent of the big data era, multi-class learning problems-such as text or image classification - can involve tens or hundreds of thousands of classes, and the samples are usually spatially-connected or temporally-correlated with their physically-connected neighbours, which can not be directly handled by these existing methods or algorithms. This implies that many samples are not satisfy the i.i.d. assumption or this assumption is very restrictive and can not be strictly justified in real-world problems. In addition, the AIO methods above are theoretically reasonable (i.e., Fisher consistent) and experiments indicate that they can lead to well-generalizing hypothesis, but efficient algorithms are lacking so far. Then three problems are posed: • Whether the AIO-MSVM algorithm based on non-i.i.d. samples is consistent? If this algorithm is consistent, what is the learning rate of this algorithm with non-i.i.d. samples? Whether the efficient AIO-MSVM algorithm can be proposed (or whether the AIO-MSVM algorithm with i.i.d. samples can be improved)? • The goal of this paper is to answer these problems above. Now, let' s highlight the main features of this paper. • The generalization bound of AIO-MSVM algorithm with u.e.M.c. samples is established and the fast learning rate is obtained. The AIO-MSVM algorithm with u.e.M.c. samples is proved to be consistent. To our knowledge, these studies are the first works on this topic. • A new algorithm, called AIO-MSVM algorithm based on q -times Markovian resampling (AIO-Mar- q), is proposed. The effectiveness and efficiency of the proposed algorithm are validated by experiments on public datasets. The generalization performance of "All-in-one" Multi-class SVM (AIO-MSVM) based on uniformly ergodic Markovian chain (u.e.M.c.) samples is considered. We establish the fast learning rate of AIO-MSVM algorithm with u.e.M.c. samples and prove that AIO-MSVM algorithm with u.e.M.c. samples is consistent. We also propose a novel AIO-MSVM algorithm based on q -times Markovian resampling (AIO-MSVM-MR), and show the numerical investigation on the learning performance of AIO-MSVM-MR based on public datasets. The experimental studies indicate that compared to the classical AIO-MSVM algorithm and other MSVM algorithms, the proposed AIO-MSVM-MR algorithm has not only smaller misclassification rate, but also less sampling and training total time. We present some discussions on the case of unbalanced training samples, the choices of q and two technical parameters, and present some explanations on the learning performance of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Active learning alleviates labeling costs by selecting and labeling the most informative examples from an unlabeled pool. However, most existing active learning approaches estimate informativeness with uncalibrated confidence, resulting in unreliable informativeness estimation. These approaches generally ignored two significant issues caused by uncalibrated confidence methods. Firstly, the average uncalibrated confidence generated by modern neural networks is usually higher than the accuracy. Secondly, examples located near the decision boundaries are unstable during prediction when the target model updates parameters in the last several epochs, even throughout the training process. This phenomenon, caused by the forgetting characteristic of neural networks, has a significant impact on some specific models that estimate the informativeness by predicted probability vectors or pseudo labels. To address these issues, in this paper, we propose a novel active learning approach to reliably estimate informativeness with calibrated confidence. Specifically, we integrate the intermediate predictions for each unlabeled example, generated by the target model during the training process, to generate calibrated confidence. The calibrated confidence can capture a tendentious label from an indecisive subset of the class space. We show that the calibrated confidence with tendentiousness can maintain the ability of correct predictions. The empirical results demonstrate that our approach outperforms the state-of-the-art active learning methods on image classification tasks. • The active learning algorithm selects unstable examples by calibrated confidence.. • The confidence calibration method only requires training a single model. • Extensive experiments show the competitive performance of the proposed approach. [ABSTRACT FROM AUTHOR]

Medical data sharing across institutes is crucial to large-scale multi-center studies and the development of real-world AI applications but suffers from serious privacy issues. A promising solution to address this challenge is federated learning, which typically aggregates a global model from heterogeneous data spread across numerous clients without exchanging data. However, the traditional federated learning algorithm (i.e. , FedAvg) merely aggregates the locally distributed models according to the amount of data on each client and lacks the consideration of data heterogeneity. In this paper, we propose a novel Federated Adaptive Reweighting (FedAR) algorithm for medical image classification. FedAR employs a flexible re-weighting scheme that can balance adaptively the contributions of the amount of data and the performance of the local model on each client to the weight of that client. Specifically, we allow the amount of local data to contribute more to the weight of each client in the early training stage and let the performance of the local model play a more important role in the late stage. We have evaluated the proposed FedAR algorithm against the locally trained model, globally trained baseline, and two existing federated learning algorithms on the ISIC2018 dataset and Chest X-ray14 dataset under the settings with a variable number of clients. Our results suggest that FedAR is an effective federated learning algorithm that substantially outperforms existing federated learning approaches. • Propose FedAR, based on federated learning, for medical image classification. • Develop an adaptive reweighting scheme to address the heterogeneity issue. • Demonstrate the effectiveness of FedAR on the ISIC2018 and Chest X-ray14 datasets. • FedAR improves the federated learning with limited extra communication costs. [ABSTRACT FROM AUTHOR]

• We propose a federated contrastive learning method, which uses the feature representation learned by the global model and the local model for contrastive learning, so that the final model can gradually approach the global model in the training process to obtain better model generalization. • In view of the current problem of data heterogeneity in real medical scenes, our method effectively reduces its impact on experimental results, that is, it can show high performance and generalization in the scenario of data heterogeneity. • We conducted experiments on two medical image classification datasets, i.e., skin lesion classification dataset for melanoma and chest X-ray dataset for detecting COVID-19. Extensive experiments demonstrate the efficacy of our approach. Federated learning, which allows distributed medical institutions to train a shared deep learning model with privacy protection, has become increasingly popular recently. However, in practical application, due to data heterogeneity between different hospitals, the performance of the model will be degraded in the training process. In this paper, we propose a federated contrastive learning (FedCL) approach. FedCL integrates the idea of contrastive learning into the federated learning framework. Specifically, it combines the local model and the global model for contrastive learning, so that the local model gradually approaches the global model with the increase of communication rounds, which improves the generalization ability of the model. We validate our method on two public datasets. Extensive experiments show that our method is superior to other federated learning algorithms in medical image classification. [ABSTRACT FROM AUTHOR]

Motivated by researching new image texture modeling that improves state-of-the-art LBP variants and non-LBP descriptors, this paper proposes a novel approach for constructing local image descriptors, which are suitable for histogram based image representation. Instead of heuristic code constructions, the proposed approach is based on local concave-and-convex characteristics, which have high ability to extract discriminative and stable texture representation. Different from the majority of descriptors that only encode relationships between the pixels in doublets around central pixel (within 3  ×  3 neighborhood), the proposed approach encodes relationships between the pixels in triplets by including the central pixel in the modeling. We build two distinct descriptors by dividing local features into two distinct groups, i.e., local concave and convex microstructure patterns (LCvMSP and LCxMSP), according to relationships between the pixels inside the triplets, formed along closed path around the central pixel of a 3 × 3-grayscale image patch. To make the descriptors more insensitive to noise and invariant to monotonic gray scale transformation, two supplementary triplets are added in the modeling. These triplets are formed using the central pixel and four virtual pixels set to the median of the grey-scale values of the 3 × 3 neighbourhood and the whole image and the average local and global gray levels respectively. The histograms obtained from the single scale descriptors LCvMSP and LCxMSP are concatenated together to build multi-scale histogram feature vector referred to as local concave-and-convex micro-structure pattern (LCCMSP), that is expected to better represent salient local texture structure. We evaluated the effectiveness of the proposed methods on thirteen challenging representative widely-used texture datasets, and found that the proposed LCvMSP, LCxMSP and LCCMSP operators achieve performances that are competitive or better than a large number of recent most promising state-of- the-art LBP variants and non-LBP descriptors. Statistical comparison based on Wilcoxon signed rank test demonstrated that the proposed methods are the top three over all the tested datasets. [ABSTRACT FROM AUTHOR]

Nonlinear methods such as Deep Neural Networks (DNNs) are the gold standard for various challenging machine learning problems such as image recognition. Although these methods perform impressively well, they have a significant disadvantage, the lack of transparency, limiting the interpretability of the solution and thus the scope of application in practice. Especially DNNs act as black boxes due to their multilayer nonlinear structure. In this paper we introduce a novel methodology for interpreting generic multilayer neural networks by decomposing the network classification decision into contributions of its input elements. Although our focus is on image classification, the method is applicable to a broad set of input data, learning tasks and network architectures. Our method called deep Taylor decomposition efficiently utilizes the structure of the network by backpropagating the explanations from the output to the input layer. We evaluate the proposed method empirically on the MNIST and ILSVRC data sets. [ABSTRACT FROM AUTHOR]

Multiple-instance learning (MIL) has been a popular topic in the study of pattern recognition for years due to its usefulness for such tasks as drug activity prediction and image/text classification. In a typical MIL setting, a bag contains a bag-level label and more than one instance/pattern. How to bridge instance-level representations to bag-level labels is a key step to achieve satisfactory classification accuracy results. In this paper, we present a supervised learning method, diversified dictionaries MIL, to address this problem. Our approach, on the one hand, exploits bag-level label information for training class-specific dictionaries. On the other hand, it introduces a diversity regularizer into the class-specific dictionaries to avoid ambiguity between them. To the best of our knowledge, this is the first time that the diversity prior is introduced to solve the MIL problems. Experiments conducted on several benchmark (drug activity and image/text annotation) datasets show that the proposed method compares favorably to state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In sparse dictionary learning based face recognition (FR), a discriminative dictionary is learned from the training set so that good classification performance can be achieved on probe set. In order to achieve better performance and less computation, dimensionality reduction is applied on source data before training. Most of the proposed dictionary learning methods learn features and dictionary separatively, which may decrease the power of dictionary learning because the classification ability of dictionary learning method is based on data structure of source domain. Therefore, a sparse embedded dictionary learning method (SEDL) is proposed, of which dictionary learning and dimensionality reduction are jointly realized and the margin of coefficients distance between between-class and within-class is encourage to be large in order to enhance the classification ability and gain discriminative information. Moreover, orthogonality of the projection matrix is preserved which is critical to data reconstruction. And data reconstruction is considered to be important for sparse representation. In this paper, an extension of discriminant dictionary learning and sparse embedding is proposed and realized with novel strategies. Experiments show that our method achieves better performance than other state-of-art methods on face recognition. [ABSTRACT FROM AUTHOR]

The human visual system proves expert in discovering patterns in both global and local feature space. Can we design a similar way for unsupervised feature learning? In this paper, we propose a novel spatial pooling method within an unsupervised feature learning framework, named Rich and Robust Feature Pooling ( R 2 FP ), to better extract rich and robust representation from sparse feature maps learned from the raw data. Both local and global pooling strategies are further considered to instantiate such a method. The former selects the most representative features in the sub-region and summarizes the joint distribution of the selected features, while the latter is utilized to extract multiple resolutions of features and fuse the features with a feature balance kernel for rich representation. Extensive experiments on several image recognition tasks demonstrate the superiority of the proposed method. [ABSTRACT FROM AUTHOR]

Previous works have demonstrated that image classification performance can be significantly improved by manifold learning. However, performance of manifold learning heavily depends on the manual selection of parameters, resulting in bad adaptability in real-world applications. In this paper, we propose a new dimensionality reduction method called discriminative sparsity preserving projections (DSPP). Different from the existing sparse subspace algorithms, which manually construct a penalty adjacency graph, DSPP employs sparse representation model to adaptively build both intrinsic adjacency graph and penalty graph with weight matrix, and then integrates global within-class structure into the discriminant manifold learning objective function for dimensionality reduction. Extensive experimental results on four image databases demonstrate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

A novel approach for automatic segmentation and recognition of CAPTCHAs with variable orientation and random collapse of overlapped characters is presented in this paper. Additionally, the extension of the proposed approach to break reCAPTCHA of version of 2012 is also discussed. The original proposal consists in straightening characters and word in CAPTCHA exploiting then a three-color bar code for their segmentation. The recognition of straightened characters and whole word is provided by the proposed original SVM-based learning classifier. The main goal of this research is to reduce vulnerability of CAPTCHA from spam and frauds as well as to provide an approach for recognizing either handwritten or degraded and damaged texts in ancient manuscripts by OCR systems. The designed framework for breaking CAPTCHAs by the proposed approach has been tested achieving average segmentation success rate up to 82% for reCAPTCHA of version 2011 and achieving 95.5% by extended approach for reCAPTCHA of version 2012 with response time less than 0.5 s per two-word reCAPTCHA. The implemented SVM classifier shows a competitive precision about 94%. The obtained very satisfactory results confirm that the proposed approach may be used for development of new security mechanisms to protect users against cyber-criminal activities and Internet threats. [ABSTRACT FROM AUTHOR]

In this paper, we introduce a novel framework for text identification and recognition, called tensor representation learning based image patch analysis (TRL-IPA). Unlike most of previous text identification approaches, which can only be applied to binarized images, TRL-IPA can be directly applied to gray level and color images. TRL-IPA is built on a general formulation of the convergent tensor representation learning (CTRL) algorithms. In the implementation of TRL-IPA, image patches are represented in the form of tensors, while low dimensional representations of these tensors are learned via a CTRL algorithm. To identify text regions in new coming document images, a random forest classifier is trained in the learned tensor subspace. Moreover, the TRL-IPA framework can be straightforwardly applied to recognition problems, such as handwritten digits recognition. We conducted extensive experiments on ancient Chinese, Arabic and Cyrillic document images, to evaluate TRL-IPA on text identification tasks. Experimental results demonstrate its effectiveness and robustness. In addition, recognition results on images of handwritten digits show its advantage over state-of-the-art vector and tensor representation based approaches. [ABSTRACT FROM AUTHOR]

• Learn class-specific discriminative distance metrics for different classes. • Incorporate least squares regression to relax the class-specific metric learning. • Extensive experiments demonstrate the effectiveness of our method. Metric learning aims to learn an appropriate distance metric for a given machine learning task. Despite its impressive performance in the field of image recognition, it may still not be discriminative enough for scene recognition because of the high within-class diversity and high between-class similarity of scene images. In this paper, we propose a novel class-specific discriminative metric learning method (CSDML) to alleviate these problems. More specifically, we learn a distinctive linear transformation for each class (or, equivalently, a Mahalanobis distance metric for each class), which allows to project the samples of that class into a corresponding low-dimensional discriminative space. The overall aim is to simultaneously minimize the Euclidean distances between the projections of samples of the same class (or, equivalently, the Mahalanobis distances between these samples) and maximize the Euclidean distances between the projections of samples of different classes. Additionally, we incorporate least squares regression into the optimization problem, rendering class-specific metric learning more flexible and better suited to tackle scene recognition. Experimental results on four benchmark scene datasets demonstrate that the proposed method outperforms most of the state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

• A new Chinese-Herbs dataset is constructed, supporting the research of Chinese herbal recognition and machine learning. • A new Attentional Pyramid Networks is proposed, where competitive attention and spatial collaborative attention are proposed. • The proposed CA selectively emphasizes the features of different level whereas SCA quantifies the importance of each region. • A framework with APN for Chinese herbal recognition is proposed. Experiment results verify the effectiveness of our methods. Chinese herbs play a critical role in Traditional Chinese Medicine. Due to different recognition granularity, they can be recognized accurately only by professionals with much experience. It is expected that they can be recognized automatically using new techniques like machine learning. However, there is no Chinese herbal image dataset available. Simultaneously, there is no machine learning method which can deal with Chinese herbal image recognition well. Therefore, this paper begins with building a new standard Chinese-Herbs dataset. Subsequently, a new Attentional Pyramid Networks (APN) for Chinese herbal recognition is proposed, where both novel competitive attention and spatial collaborative attention are proposed and then applied. APN can adaptively model Chinese herbal images with different feature scales. Finally, a new framework for Chinese herbal recognition is proposed as a new application of APN. Experiments are conducted on our constructed dataset and validate the effectiveness of our methods. [ABSTRACT FROM AUTHOR]