Haijun Lei

Reversible watermarking scheme for medical image based on differential evolution

A reversible watermarking method is proposed with wavelet transforms and SVD.Signature and logo data are inserted by recursive dither modulation algorithm.DE is explored to design the quantization steps optimally.Good balance of imperceptibility, robustness and capacity is obtained by DE.Experiments show good performance and outperform the related algorithms. Currently, most medical images are stored and exchanged with little or no security; hence it is important to provide protection for the intellectual property of these images in a secured environment. In this paper, a new and reversible watermarking method is proposed to address this security issue. Specifically, signature information and textual data are inserted into the original medical images based on recursive dither modulation (RDM) algorithm after wavelet transform and singular value decomposition (SVD). In addition, differential evolution (DE) is applied to design the quantization steps (QSs) optimally for controlling the strength of the watermark. Using these specially designed hybrid techniques, the proposed watermarking technique obtains good imperceptibility and high robustness. Experimental results indicate that the proposed method is not only highly competitive, but also outperforms the existing methods.

Optimal and secure audio watermarking scheme based on self-adaptive particle swarm optimization and quaternion wavelet transform

In this paper, a new audio watermarking scheme based on self-adaptive particle swarm optimization (SAPSO) and quaternion wavelet transform (QWT) is proposed. By obtaining optimal watermark strength using a uniquely designed objective function, SAPSO addresses the conflicting problem of robustness, imperceptibility, and capacity of audio watermarking scheme using self-adjusted parameters. To withstand de-synchronization attack, a synchronization sequence generated by chaotic signals is also adopted in our scheme. Furthermore, the utilization of chaotic signals significantly enhances the security of the proposed scheme. The experimental results validate that our scheme is not only robust against de-synchronization attack, but also typical signal manipulations and StirMark attack. Our comparative analysis also revealed that the proposed scheme outperforms the state-of-the-arts audio watermarking schemes. A new audio watermarking scheme based on quaternion wavelet transform is proposed.Self-adaptive particle swarm optimization is developed to optimize the parameters.Synchronization code is inserted to withstand de-synchronization attacks.The security is enhanced by chaotic maps.The proposed method is very robust to resampling and cropping attacks.

Robust watermarking scheme for medical image using optimization method

Currently, most medical images are stored and exchanged without any consideration of security, and hence it is highly desirable to provide content protection for medical images that are used in a secured environment. In this paper, a new and robust watermarking method is proposed to address this security issue. Specifically, signature information and textual data are inserted into the original medical images. The integer wavelet transform (IWT) are combined with singular value decomposition (SVD) to provide robustness in the proposed method. Meanwhile, differential evolution (DE) is used to optimally design the quantization steps (QSs) for controlling the watermarking strength. With hybrid techniques such as IWT, SVD, and DE, the proposed watermarking algorithm not only achieves good imperceptibility, but also high robustness. Experimental results demonstrate that the proposed method outperforms existing methods.

A deeply supervised residual network for HEp-2 cell classification via cross-modal transfer learning

Abstract Accurate Human Epithelial-2 (HEp-2) cell image classification plays an important role in the diagnosis of many autoimmune diseases and subsequent treatment. One of the key challenges is huge intra-class variations caused by inhomogeneous illumination. To address it, we propose a framework based on very deep supervised residual network (DSRN) to classify HEp-2 cell images. Specifically, we adopt a residual network of 50 layers (ResNet-50) that is substantially deep to extract rich and discriminative features. The deep supervision is imposed on the ResNet-based framework to further boost the classification performance by directly guiding the training of the lower and upper levels of the network. The proposed method is evaluated using two publicly available datasets (i.e., International Conference on Pattern Recognition (ICPR) 2012 and ICPR2016-Task1 cell classification contest datasets). Different from the previous deep learning models learned from scratch, a cross-modal transfer learning strategy is developed. Namely, we pretrain ICPR2012 dataset to fine-tune ICPR2016 dataset based on our DSRN model since both datasets are similar. Extensive experiments show that the proposed method delivers state-of-the-art performance and outperforms the traditional methods based on deep convolutional neural network (DCNN).

Multipurpose watermarking scheme via intelligent method and chaotic map

In this paper, we propose a novel multipurpose intelligent image watermarking scheme for both content authentication and copyright protection. To achieve this, we first utilize integer discrete wavelet transform (IDWT) for watermark insertion and detection. The low frequency component of IDWT is used to insert a robust watermark (e.g., the copyright information) for copyright protection, whereas the high frequency component of IDWT is used to embed a fragile watermark (e.g., the logo data) for content authentication. To achieve a good tradeoff among the watermark conflicting requirements (e.g., robustness, fidelity and capacity), we develop an artificial bee colony (ABC) algorithm for optimal parameter selection. Experimental results demonstrate that the proposed scheme achieves promising performance in both copyright protection and content authentication simultaneously, which confirm the superiority of our proposed algorithm as compared to existing methods.

Object recognition based on adapative bag of feature and discriminative learning

In this paper, a new method is proposed to incorporate the saliency map to weight the extracted features with discriminative technique for learning the spatial discriminative information of images. Different from the conventional bag of word (BoW) approach, the descriptive bag of phrase approach is explored to capture the word co-occurrence and dependence. The image score based on the saliency map is learned to optimize the support vector machine (SVM) parameter. Discriminative learning techniques are adopted based on image score and fed into the SVM classifier. Moreover, the histogram intersection mapping and normalization method is further adopted to enhance the classification performance. Experimental results on the 3 popular databases demonstrate the effectiveness of the method and show the promising performance over the existing state-of-the-art methods.

Longitudinal and Multi-modal Data Learning via Joint Embedding and Sparse Regression for Parkinson’s Disease Diagnosis

Parkinson’s disease (PD) is a neurodegenerative progressive disease that mainly affects the motor systems of patients. To slow this disease deterioration, robust and accurate diagnosis of PD is an effective way to alleviate mental and physical sufferings of clinical intervention. In this paper, we propose a new unsupervised feature selection method via joint embedding learning and sparse regression using longitudinal multi-modal neuroimaging data. Specifically, the proposed method performs feature selection and local structure learning, simultaneously, to adaptively determine the similarity matrix. Meanwhile, we constrain the similarity matrix to make it contains c connected components for gaining the most accurate information of the neuroimaging data structure. The baseline data is utilized to establish the feature selection model to select the most discriminative features. Namely, we exploit baseline data to train four regression models for the clinical scores prediction (depression, sleep, olfaction, and cognition scores) and a classification model for the classification of PD disease in the future time point. Extensive experiments are conducted to demonstrate the effectiveness of the proposed method on the Parkinson’s Progression Markers Initiative (PPMI) dataset. The experimental results demonstrate that, our proposed method can enhance the performance in clinical scores prediction and class label identification in longitudinal data and outperforms the state-of-art methods as well.

Multi-task Sparse Low-Rank Learning for Multi-classification of Parkinson's Disease.

Identifying prodromal stages of Parkinson’s disease (PD) draws increasing recognition as non-motor symptoms may appear before classical clinical diagnosis based on motor signs. To effectively develop a computer-aided diagnosis for multiple disease progression stages, neuroimaging has been widely applied for its convenience of revealing the intricate brain structure. However, the high dimensional neuroimaging features and limited sample size bring the main challenges for the diagnosis task. To handle it, a multi-task sparse low-rank learning framework is proposed to unveil the underlying relationships between input data and output targets by building a matrix-regularized feature network. Inductions of multiple tasks are simultaneously performed to capture intrinsic feature relatedness with multi-task learning. By discarding the irrelevant features and preserving the discriminative structured features, our proposed method can select the most relevant features and identify different stages of PD with different multi-classification models. Extensive experimental results on the Parkinson’s progression markers initiative (PPMI) dataset demonstrate that the proposed method achieves promising classification performance and outperforms the conventional algorithms.

Joint detection and clinical score prediction in Parkinson's disease via multi-modal sparse learning

In this study, a novel feature selection framework is proposed to simultaneously perform classification and clinical scores prediction of Parkinsons disease (PD) via multi-modal neuroimaging data. Specifically, a new feature selection model is devised to capture discriminative features to train support vector regression model for clinical scores (e.g., sleep scores and olfactory scores) prediction and support vector classification model for class label identification. Our method is evaluated on a public dataset of 208 subjects including 56 normal controls (NC), 123 PD and 29 scans without evidence of dopamine deficit (SWEDD) via a 10-fold cross-validation method. The experimental results demonstrate that multi-modal data can effectively improve the performance in disease status identification and clinical scores prediction compared to one single modality. Our proposed method also outperforms the related methods.

Multipurpose and intelligent watermarking scheme for medical data

In this paper, a new multipurpose watermarking scheme is proposed to provide protection and authentication of medical data. To achieve both purposes, integer discrete wavelet transform (IDWT) decomposition is first performed, and then robust watermark (e.g. logo data) is embedded in the low frequency IDWT sub-band for copyright protection, while fragile watermark (e.g. the diagnosis information) is inserted in the high frequency IDWT sub-band for tampering detection. A tradeoff between conflicting watermark requirements such as robustness, capacity and imperceptibility is achieved by particle swarm optimization (PSO) training technique. Experimental results validate the effectiveness and efficiency of the proposed algorithm. The achieved fragileness and robustness confirmed that the proposed scheme has capability of rightful ownership protection and authentication simultaneously.