arXiv subjects

Shanshan Wang

Publications and source records attributed to Shanshan Wang.

At least 199 records · Page 11Linked to original sources

arXiv2019

Model-based Convolutional De-Aliasing Network Learning for Parallel MR Imaging

Parallel imaging has been an essential technique to accelerate MR imaging. Nevertheless, the acceleration rate is still limited due to the ill-condition and challenges associated with the undersampled reconstruction. In this paper, we propose a model-based convolutional de-aliasing network with adaptive parameter learning to achieve accurate reconstruction from multi-coil undersampled k-space data. Three main contributions have been made: a de-aliasing reconstruction model was proposed to accelerate parallel MR imaging with deep learning exploring both spatial redundancy and multi-coil correlations; a split Bregman iteration algorithm was developed to solve the model efficiently; and unlike most existing parallel imaging methods which rely on the accuracy of the estimated multi-coil sensitivity, the proposed method can perform parallel reconstruction from undersampled data without explicit sensitivity calculation. Evaluations were conducted on \emph{in vivo} brain dataset with a variety of undersampling patterns and different acceleration factors. Our results demonstrated that this method could achieve superior performance in both quantitative and qualitative analysis, compared to three state-of-the-art methods.

eess.IV↗

arXiv2019

DeepcomplexMRI: Exploiting deep residual network for fast parallel MR imaging with complex convolution

This paper proposes a multi-channel image reconstruction method, named DeepcomplexMRI, to accelerate parallel MR imaging with residual complex convolutional neural network. Different from most existing works which rely on the utilization of the coil sensitivities or prior information of predefined transforms, DeepcomplexMRI takes advantage of the availability of a large number of existing multi-channel groudtruth images and uses them as labeled data to train the deep residual convolutional neural network offline. In particular, a complex convolutional network is proposed to take into account the correlation between the real and imaginary parts of MR images. In addition, the k space data consistency is further enforced repeatedly in between layers of the network. The evaluations on in vivo datasets show that the proposed method has the capability to recover the desired multi-channel images. Its comparison with state-of-the-art method also demonstrates that the proposed method can reconstruct the desired MR images more accurately.

eess.IV↗

arXiv2019

CLCI-Net: Cross-Level fusion and Context Inference Networks for Lesion Segmentation of Chronic Stroke

Segmenting stroke lesions from T1-weighted MR images is of great value for large-scale stroke rehabilitation neuroimaging analyses. Nevertheless, there are great challenges with this task, such as large range of stroke lesion scales and the tissue intensity similarity. The famous encoder-decoder convolutional neural network, which although has made great achievements in medical image segmentation areas, may fail to address these challenges due to the insufficient uses of multi-scale features and context information. To address these challenges, this paper proposes a Cross-Level fusion and Context Inference Network (CLCI-Net) for the chronic stroke lesion segmentation from T1-weighted MR images. Specifically, a Cross-Level feature Fusion (CLF) strategy was developed to make full use of different scale features across different levels; Extending Atrous Spatial Pyramid Pooling (ASPP) with CLF, we have enriched multi-scale features to handle the different lesion sizes; In addition, convolutional long short-term memory (ConvLSTM) is employed to infer context information and thus capture fine structures to address the intensity similarity issue. The proposed approach was evaluated on an open-source dataset, the Anatomical Tracings of Lesions After Stroke (ATLAS) with the results showing that our network outperforms five state-of-the-art methods. We make our code and models available at https://github.com/YH0517/CLCI_Net.

eess.IV↗

arXiv2019

Grasping asymmetric information in market impacts

The price impact for a single trade is estimated by the immediate response on an event time scale, i.e., the immediate change of midpoint prices before and after a trade. We work out the price impacts across a correlated financial market. We quantify the asymmetries of the distributions and of the market structures of cross-impacts, and find that the impacts across the market are asymmetric and non-random. Using spectral statistics and Shannon entropy, we visualize the asymmetric information in price impacts. Also, we introduce an entropy of impacts to estimate the randomness between stocks. We show that the useful information is encoded in the impacts corresponding to small entropy. The stocks with large number of trades are more likely to impact others, while the less traded stocks have higher probability to be impacted by others.

q-fin.TR↗

arXiv2019

Manifold Criterion Guided Transfer Learning via Intermediate Domain Generation

In many practical transfer learning scenarios, the feature distribution is different across the source and target domains (i.e. non-i.i.d.). Maximum mean discrepancy (MMD), as a domain discrepancy metric, has achieved promising performance in unsupervised domain adaptation (DA). We argue that MMD-based DA methods ignore the data locality structure, which, to some extent, would cause the negative transfer effect. The locality plays an important role in minimizing the nonlinear local domain discrepancy underlying the marginal distributions. For better exploiting the domain locality, a novel local generative discrepancy metric (LGDM) based intermediate domain generation learning called Manifold Criterion guided Transfer Learning (MCTL) is proposed in this paper. The merits of the proposed MCTL are four-fold: 1) the concept of manifold criterion (MC) is first proposed as a measure validating the distribution matching across domains, and domain adaptation is achieved if the MC is satisfied; 2) the proposed MC can well guide the generation of the intermediate domain sharing similar distribution with the target domain, by minimizing the local domain discrepancy; 3) a global generative discrepancy metric (GGDM) is presented, such that both the global and local discrepancy can be effectively and positively reduced; 4) a simplified version of MCTL called MCTL-S is presented under a perfect domain generation assumption for more generic learning scenario. Experiments on a number of benchmark visual transfer tasks demonstrate the superiority of the proposed manifold criterion guided generative transfer method, by comparing with other state-of-the-art methods. The source code is available in https://github.com/wangshanshanCQU/MCTL.

cs.CV↗

arXiv2019

Low-Latency Deep Clustering For Speech Separation

This paper proposes a low algorithmic latency adaptation of the deep clustering approach to speaker-independent speech separation. It consists of three parts: a) the usage of long-short-term-memory (LSTM) networks instead of their bidirectional variant used in the original work, b) using a short synthesis window (here 8 ms) required for low-latency operation, and, c) using a buffer in the beginning of audio mixture to estimate cluster centres corresponding to constituent speakers which are then utilized to separate speakers within the rest of the signal. The buffer duration would serve as an initialization phase after which the system is capable of operating with 8 ms algorithmic latency. We evaluate our proposed approach on two-speaker mixtures from the Wall Street Journal (WSJ0) corpus. We observe that the use of LSTM yields around one dB lower SDR as compared to the baseline bidirectional LSTM in terms of source to distortion ratio (SDR). Moreover, using an 8 ms synthesis window instead of 32 ms degrades the separation performance by around 2.1 dB as compared to the baseline. Finally, we also report separation performance with different buffer durations noting that separation can be achieved even for buffer duration as low as 300ms.

cs.SD↗

arXiv2019

CRDN: Cascaded Residual Dense Networks for Dynamic MR Imaging with Edge-enhanced Loss Constraint

Dynamic magnetic resonance (MR) imaging has generated great research interest, as it can provide both spatial and temporal information for clinical diagnosis. However, slow imaging speed or long scanning time is still one of the challenges for dynamic MR imaging. Most existing methods reconstruct Dynamic MR images from incomplete k-space data under the guidance of compressed sensing (CS) or low rank theory, which suffer from long iterative reconstruction time. Recently, deep learning has shown great potential in accelerating dynamic MR. Our previous work proposed a dynamic MR imaging method with both k-space and spatial prior knowledge integrated via multi-supervised network training. Nevertheless, there was still a certain degree of smooth in the reconstructed images at high acceleration factors. In this work, we propose cascaded residual dense networks for dynamic MR imaging with edge-enhance loss constraint, dubbed as CRDN. Specifically, the cascaded residual dense networks fully exploit the hierarchical features from all the convolutional layers with both local and global feature fusion. We further utilize the total variation (TV) loss function, which has the edge enhancement properties, for training the networks.

cs.CV↗

arXiv2018

A Flexible Spatial Autoregressive Modelling Framework for Mixed Covariates of Multiple Data Types

Mixed spatial autoregressive (SAR) models with numerical covariates have been well studied. However, as non-numerical data, such as functional data and compositional data, receive substantial amounts of attention and are applied to economics, medicine and meteorology, it becomes necessary to develop flexible SAR models with multiple data types. In this article, we integrate three types of covariates, functional, compositional and numerical, in an SAR model. The new model has the merits of classical functional linear models and compositional linear models with scalar responses. Moreover, we develop an estimation method for the proposed model, which is based on functional principal component analysis (FPCA), the isometric logratio (ilr) transformation and the maximum likelihood estimation method. Monte Carlo experiments demonstrate the effectiveness of the estimators. A real dataset is also used to illustrate the utility of the proposed model.

stat.AP↗

arXiv2018

DIMENSION: Dynamic MR Imaging with Both K-space and Spatial Prior Knowledge Obtained via Multi-Supervised Network Training

Dynamic MR image reconstruction from incomplete k-space data has generated great research interest due to its capability in reducing scan time. Nevertheless, the reconstruction problem is still challenging due to its ill-posed nature. Most existing methods either suffer from long iterative reconstruction time or explore limited prior knowledge. This paper proposes a dynamic MR imaging method with both k-space and spatial prior knowledge integrated via multi-supervised network training, dubbed as DIMENSION. Specifically, the DIMENSION architecture consists of a frequential prior network for updating the k-space with its network prediction and a spatial prior network for capturing image structures and details. Furthermore, a multisupervised network training technique is developed to constrain the frequency domain information and reconstruction results at different levels. The comparisons with classical k-t FOCUSS, k-t SLR, L+S and the state-of-the-art CNN-based method on in vivo datasets show our method can achieve improved reconstruction results in shorter time.

cs.CV↗

arXiv2018

Spatial Functional Linear Model and its Estimation Method

The classical functional linear regression model (FLM) and its extensions, which are based on the assumption that all individuals are mutually independent, have been well studied and are used by many researchers. This independence assumption is sometimes violated in practice, especially when data with a network structure are collected in scientific disciplines including marketing, sociology and spatial economics. However, relatively few studies have examined the applications of FLM to data with network structures. We propose a novel spatial functional linear model (SFLM), that incorporates a spatial autoregressive parameter and a spatial weight matrix into FLM to accommodate spatial dependencies among individuals. The proposed model is relatively flexible as it takes advantage of FLM in handling high-dimensional covariates and spatial autoregressive (SAR) model in capturing network dependencies. We develop an estimation method based on functional principal component analysis (FPCA) and maximum likelihood estimation. Simulation studies show that our method performs as well as the FPCA-based method used with FLM when no network structure is present, and outperforms the latter when network structure is present. A real weather data is also employed to demonstrate the utility of the SFLM.

stat.CO↗

arXiv2018

Classifying Mammographic Breast Density by Residual Learning

Mammographic breast density, a parameter used to describe the proportion of breast tissue fibrosis, is widely adopted as an evaluation characteristic of the likelihood of breast cancer incidence. In this study, we present a radiomics approach based on residual learning for the classification of mammographic breast densities. Our method possesses several encouraging properties such as being almost fully automatic, possessing big model capacity and flexibility. It can obtain outstanding classification results without the necessity of result compensation using mammographs taken from different views. The proposed method was instantiated with the INbreast dataset and classification accuracies of 92.6% and 96.8% were obtained for the four BI-RADS (Breast Imaging and Reporting Data System) category task and the two BI-RADS category task,respectively. The superior performances achieved compared to the existing state-of-the-art methods along with its encouraging properties indicate that our method has a great potential to be applied as a computer-aided diagnosis tool.

cs.CV↗

arXiv2018

Fine-grained visual recognition with salient feature detection

Computer vision based fine-grained recognition has received great attention in recent years. Existing works focus on discriminative part localization and feature learning. In this paper, to improve the performance of fine-grained recognition, we try to precisely locate as many salient parts of object as possible at first. Then, we figure out the classification probability that can be obtained by using separate parts for object classification. Finally, through extracting efficient features from each part and combining them, then feeding to a classifier for recognition, an improved accuracy over state-of-art algorithms has been obtained on CUB200-2011 bird dataset.

cs.CV↗

arXiv2018

Object Activity Scene Description, Construction and Recognition

Action recognition is a critical task for social robots to meaningfully engage with their environment. 3D human skeleton-based action recognition is an attractive research area in recent years. Although, the existing approaches are good at action recognition, it is a great challenge to recognize a group of actions in an activity scene. To tackle this problem, at first, we partition the scene into several primitive actions (PAs) based upon motion attention mechanism. Then, the primitive actions are described by the trajectory vectors of corresponding joints. After that, motivated by text classification based on word embedding, we employ convolution neural network (CNN) to recognize activity scenes by considering motion of joints as "word" of activity. The experimental results on the scenes of human activity dataset show the efficiency of the proposed approach.

cs.CV↗

arXiv2018

Local fluctuations of the signed traded volumes and the dependencies of demands: a copula analysis

We investigate how the local fluctuations of the signed traded volumes affect the dependence of demands between stocks. We analyze the empirical dependence of demands using copulas and show that they are well described by a bivariate $\mathcal{K}$ copula density function. We find that large local fluctuations strongly increase the positive dependence but lower slightly the negative one in the copula density. This interesting feature is due to cross-correlations of volume imbalances between stocks. Also, we explore the asymmetries of tail dependencies of the copula density, which are moderate for the negative dependencies but strong for the positive ones. For the latter, we reveal that large local fluctuations of the signed traded volumes trigger stronger dependencies of demands than of supplies, probably indicating a bull market with persistent raising of prices.

q-fin.TR↗

arXiv2017

Statistical properties of market collective responses

We empirically analyze the price and liquidity responses to trade signs, traded volumes and signed traded volumes. Utilizing the singular value decomposition, we explore the interconnections of price responses and of liquidity responses across the whole market. The statistical characteristics of their singular vectors are well described by the $t$ location-scale distribution. Furthermore, we discuss the relation between prices and liquidity with respect to their overlapping factors. The factors of price and liquidity changes are non-random when these factors are related to the traded volumes. This means that the traded volumes play a critical role in the price change induced by the liquidity change. In contrast, the two kinds of factors are weakly overlapping when they are related to the trade signs and signed traded volumes. Hence, an imbalance of liquidity is related to the price change.

q-fin.ST↗

arXiv2017

Microscopic Understanding of Cross-Responses between Stocks: a Two-Component Price Impact Model

We construct a price impact model between stocks in a correlated market. For the price change of a given stock induced by the short-run liquidity of this stock itself and of the information about other stocks, we introduce a self- and a cross-impact function of the time lag. We model the average cross-response functions for individual stocks employing the impact functions of the time lag, the impact functions of traded volumes and the trade-sign correlators. To quantify the self- and cross-impacts, we propose a construction to fix the parameters in the impact functions. These parameters are further corroborated by a diffusion function that measures the correlated motion of prices from different stocks. This construction is mainly ad hoc and alternative ones are not excluded. It turns out that both the sign cross- and self-correlators are connected with the cross-responses. The self- and cross-impact functions are indispensable to compensate amplification effects which are due to the sign correlators integrated over time. We further quantify and interpret the price impacts of time lag in terms of temporary and permanent components. To support our model, we also analyze empirical data, in particular the memory properties of the sign self- and average cross-correlators. The relation between the average cross-responses and the traded volumes which are smaller than their average is of power-law form.

q-fin.TR↗

arXiv2017

Trading strategies for stock pairs regarding to the cross-impact cost

We extend the framework of trading strategies of Gatheral [2010] from single stocks to a pair of stocks. Our trading strategy with the executions of two round-trip trades can be described by the trading rates of the paired stocks and the ratio of their trading periods. By minimizing the potential cost arising from cross-impacts, i.e., the price change of one stock due to the trades of another stock, we can find out an optimal strategy for executing a sequence of trades from different stocks. We further apply the model of the strategy to a specific case, where we quantify the cross-impacts of traded volumes and of time lag with empirical data for the computation of costs. We thus picture the influence of cross-impacts on the trading strategy.

q-fin.TR↗

arXiv2016

Average cross-responses in correlated financial market

There are non-vanishing price responses across different stocks in correlated financial markets. We further study this issue by performing different averages, which identify active and passive cross-responses. The two average cross-responses show different characteristic dependences on the time lag. The passive cross-response exhibits a shorter response period with sizeable volatilities, while the corresponding period for the active cross-response is longer. The average cross-responses for a given stock are evaluated either with respect to the whole market or to different sectors. Using the response strength, the influences of individual stocks are identified and discussed. Moreover, the various cross-responses as well as the average cross-responses are compared with the self-responses. In contrast, the short memory of trade sign cross-correlation for stock pairs, the sign cross-correlation has long memory when averaged over different pairs of stocks.

q-fin.ST↗