arXiv subjects

Qianqian Yang

Publications and source records attributed to Qianqian Yang.

At least 109 records · Page 6Linked to original sources

arXiv2022

Semantic-preserved Communication System for Highly Efficient Speech Transmission

Deep learning (DL) based semantic communication methods have been explored for the efficient transmission of images, text, and speech in recent years. In contrast to traditional wireless communication methods that focus on the transmission of abstract symbols, semantic communication approaches attempt to achieve better transmission efficiency by only sending the semantic-related information of the source data. In this paper, we consider semantic-oriented speech transmission which transmits only the semantic-relevant information over the channel for the speech recognition task, and a compact additional set of semantic-irrelevant information for the speech reconstruction task. We propose a novel end-to-end DL-based transceiver which extracts and encodes the semantic information from the input speech spectrums at the transmitter and outputs the corresponding transcriptions from the decoded semantic information at the receiver. For the speech to speech transmission, we further include a CTC alignment module that extracts a small number of additional semantic-irrelevant but speech-related information for the better reconstruction of the original speech signals at the receiver. The simulation results confirm that our proposed method outperforms current methods in terms of the accuracy of the predicted text for the speech to text transmission and the quality of the recovered speech signals for the speech to speech transmission, and significantly improves transmission efficiency. More specifically, the proposed method only sends 16% of the amount of the transmitted symbols required by the existing methods while achieving about 10% reduction in WER for the speech to text transmission. For the speech to speech transmission, it results in an even more remarkable improvement in terms of transmission efficiency with only 0.2% of the amount of the transmitted symbols required by the existing method.

eess.AS↗

arXiv2022

OTFPF: Optimal Transport-Based Feature Pyramid Fusion Network for Brain Age Estimation with 3D Overlapped ConvNeXt

Chronological age of healthy brain is able to be predicted using deep neural networks from T1-weighted magnetic resonance images (T1 MRIs), and the predicted brain age could serve as an effective biomarker for detecting aging-related diseases or disorders. In this paper, we propose an end-to-end neural network architecture, referred to as optimal transport based feature pyramid fusion (OTFPF) network, for the brain age estimation with T1 MRIs. The OTFPF consists of three types of modules: Optimal Transport based Feature Pyramid Fusion (OTFPF) module, 3D overlapped ConvNeXt (3D OL-ConvNeXt) module and fusion module. These modules strengthen the OTFPF network's understanding of each brain's semi-multimodal and multi-level feature pyramid information, and significantly improve its estimation performances. Comparing with recent state-of-the-art models, the proposed OTFPF converges faster and performs better. The experiments with 11,728 MRIs aged 3-97 years show that OTFPF network could provide accurate brain age estimation, yielding mean absolute error (MAE) of 2.097, Pearson's correlation coefficient (PCC) of 0.993 and Spearman's rank correlation coefficient (SRCC) of 0.989, between the estimated and chronological ages. Widespread quantitative experiments and ablation experiments demonstrate the superiority and rationality of OTFPF network. The codes and implement details will be released on GitHub: https://github.com/ZJU-Brain/OTFPF after final decision.

cs.CV↗

arXiv2022

A resource-efficient deep learning framework for low-dose brain PET image reconstruction and analysis

18F-fluorodeoxyglucose (18F-FDG) Positron Emission Tomography (PET) imaging usually needs a full-dose radioactive tracer to obtain satisfactory diagnostic results, which raises concerns about the potential health risks of radiation exposure, especially for pediatric patients. Reconstructing the low-dose PET (L-PET) images to the high-quality full-dose PET (F-PET) ones is an effective way that both reduces the radiation exposure and remains diagnostic accuracy. In this paper, we propose a resource-efficient deep learning framework for L-PET reconstruction and analysis, referred to as transGAN-SDAM, to generate F-PET from corresponding L-PET, and quantify the standard uptake value ratios (SUVRs) of these generated F-PET at whole brain. The transGAN-SDAM consists of two modules: a transformer-encoded Generative Adversarial Network (transGAN) and a Spatial Deformable Aggregation Module (SDAM). The transGAN generates higher quality F-PET images, and then the SDAM integrates the spatial information of a sequence of generated F-PET slices to synthesize whole-brain F-PET images. Experimental results demonstrate the superiority and rationality of our approach.

eess.IV↗

arXiv2022

Semantic-aware Speech to Text Transmission with Redundancy Removal

eess.AS↗

arXiv2022

Generalisation of continuous time random walk to anomalous diffusion MRI models with an age-related evaluation of human corpus callosum

Diffusion MRI measures of the human brain provide key insight into microstructural variations across individuals and into the impact of central nervous system diseases and disorders. One approach to extract information from diffusion signals has been to use biologically relevant analytical models to link millimetre scale diffusion MRI measures with microscale influences. The other approach has been to represent diffusion as an anomalous transport process and infer microstructural information from the different anomalous diffusion equation parameters. In this study, we investigated how parameters of various anomalous diffusion models vary with age in the human brain white matter, particularly focusing on the corpus callosum. We first unified several established anomalous diffusion models (the super-diffusion, sub-diffusion, quasi-diffusion and fractional Bloch-Torrey models) under the continuous time random walk modelling framework. This unification allows a consistent parameter fitting strategy to be applied from which meaningful model parameter comparisons can be made. We then provided a novel way to derive the diffusional kurtosis imaging (DKI) model, which is shown to be a degree two approximation of the sub-diffusion model. This link between the DKI and sub-diffusion models led to a new robust technique for generating maps of kurtosis and diffusivity using the sub-diffusion parameters \b{eta}_SUB and D_SUB. Superior tissue contrast is achieved in kurtosis maps based on the sub-diffusion model. 7T diffusion weighted MRI data for 65 healthy participants in the age range 19-78 years was used in this study. Results revealed that anomalous diffusion model parameters α and \b{eta} have shown consistent positive correlation with age in the corpus callosum, indicating α and \b{eta} are sensitive to tissue microstructural changes in aging.

physics.med-ph↗

arXiv2021

JMSNAS: Joint Model Split and Neural Architecture Search for Learning over Mobile Edge Networks

The main challenge to deploy deep neural network (DNN) over a mobile edge network is how to split the DNN model so as to match the network architecture as well as all the nodes' computation and communication capacity. This essentially involves two highly coupled procedures: model generating and model splitting. In this paper, a joint model split and neural architecture search (JMSNAS) framework is proposed to automatically generate and deploy a DNN model over a mobile edge network. Considering both the computing and communication resource constraints, a computational graph search problem is formulated to find the multi-split points of the DNN model, and then the model is trained to meet some accuracy requirements. Moreover, the trade-off between model accuracy and completion latency is achieved through the proper design of the objective function. The experiment results confirm the superiority of the proposed framework over the state-of-the-art split machine learning design methods.

cs.AI↗

arXiv2021

Blind Channel Estimation for MIMO Systems via Variational Inference

In this paper, we investigate the blind channel estimation problem for MIMO systems under Rayleigh fading channel. Conventional MIMO communication techniques require transmitting a considerable amount of training symbols as pilots in each data block to obtain the channel state information (CSI) such that the transmitted signals can be successfully recovered. However, the pilot overhead and contamination become a bottleneck for the practical application of MIMO systems with the increase of the number of antennas. To overcome this obstacle, we propose a blind channel estimation framework, where we introduce an auxiliary posterior distribution of CSI and the transmitted signals given the received signals to derive a lower bound to the intractable likelihood function of the received signal. Meanwhile, we generate this auxiliary distribution by a neural network based variational inference framework, which is trained by maximizing the lower bound. The optimal auxiliary distribution which approaches real prior distribution is then leveraged to obtain the maximum a posterior (MAP) estimation of channel matrix and transmitted data. The simulation results demonstrate that the performance of the proposed blind channel estimation method closely approaches that of the conventional pilot-aided methods in terms of the channel estimation error and symbol error rate (SER) of the detected signals even without the help of pilots.

eess.SP↗

arXiv2021

FTPipeHD: A Fault-Tolerant Pipeline-Parallel Distributed Training Framework for Heterogeneous Edge Devices

With the increased penetration and proliferation of Internet of Things (IoT) devices, there is a growing trend towards distributing the power of deep learning (DL) across edge devices rather than centralizing it in the cloud. This development enables better privacy preservation, real-time responses, and user-specific models. To deploy deep and complex models to edge devices with limited resources, model partitioning of deep neural networks (DNN) model is necessary, and has been widely studied. However, most of the existing literature only considers distributing the inference model while still relying centralized cloud infrastructure to generate this model through training. In this paper, we propose FTPipeHD, a novel DNN training framework that trains DNN models across distributed heterogeneous devices with fault tolerance mechanism. To accelerate the training with time-varying computing power of each device, we optimize the partition points dynamically according to real-time computing capacities. We also propose a novel weight redistribution approach that replicates the weights to both the neighboring nodes and the central node periodically, which combats the failure of multiple devices during training while incurring limited communication cost. Our numerical results demonstrate that FTPipeHD is 6.8x faster in training than the state of the art method when the computing capacity of the best device is 10x greater than the worst one. It is also shown that the proposed method is able to accelerate the training even with the existence of device failures.

cs.LG↗

arXiv2021

Communication-Efficient Federated Learning with Binary Neural Networks

Federated learning (FL) is a privacy-preserving machine learning setting that enables many devices to jointly train a shared global model without the need to reveal their data to a central server. However, FL involves a frequent exchange of the parameters between all the clients and the server that coordinates the training. This introduces extensive communication overhead, which can be a major bottleneck in FL with limited communication links. In this paper, we consider training the binary neural networks (BNN) in the FL setting instead of the typical real-valued neural networks to fulfill the stringent delay and efficiency requirement in wireless edge networks. We introduce a novel FL framework of training BNN, where the clients only upload the binary parameters to the server. We also propose a novel parameter updating scheme based on the Maximum Likelihood (ML) estimation that preserves the performance of the BNN even without the availability of aggregated real-valued auxiliary parameters that are usually needed during the training of the BNN. Moreover, for the first time in the literature, we theoretically derive the conditions under which the training of BNN is converging. { Numerical results show that the proposed FL framework significantly reduces the communication cost compared to the conventional neural networks with typical real-valued parameters, and the performance loss incurred by the binarization can be further compensated by a hybrid method.

cs.LG↗

arXiv2021

Sesqui-regular graphs with fixed smallest eigenvalue

Let $λ\geq2$ be an integer. For strongly regular graphs with parameters $(v, k, a,c)$ and smallest eigenvalue $-λ$, Neumaier gave two bounds on $c$ by using algebraic property of strongly regular graphs. In this paper, we will study a new class of regular graphs called sesqui-regular graphs, which contains strongly regular graphs as a subclass, and prove that for a sesqui-regular graph with parameters $(v,k,c)$ and smallest eigenvalue at least $-λ$, if $k$ is very large, then either $c \leq λ^2(λ-1)$ or $v-k-1 \leq \frac{(λ-1)^2}{4} + 1$ holds.

math.CO↗

arXiv2021

Sesqui-regular graphs with smallest eigenvalue at least $-3$

Koolen et al. showed that if a graph with smallest eigenvalue at least $-3$ has large minimal valency, then it is $2$-integrable. In this paper, we will focus on the sesqui-regular graphs with smallest eigenvalue at least $-3$ and study their integrability.

math.CO↗

arXiv2021

Fractional order magnetic resonance fingerprinting in the human cerebral cortex

Mathematical models are becoming increasingly important in magnetic resonance imaging (MRI), as they provide a mechanistic approach for making a link between tissue microstructure and signals acquired using the medical imaging instrument. The Bloch equations, which describes spin and relaxation in a magnetic field, is a set of integer order differential equations with a solution exhibiting mono-exponential behaviour in time. Parameters of the model may be estimated using a non-linear solver, or by creating a dictionary of model parameters from which MRI signals are simulated and then matched with experiment. We have previously shown the potential efficacy of a magnetic resonance fingerprinting (MRF) approach, i.e. dictionary matching based on the classical Bloch equations, for parcellating the human cerebral cortex. However, this classical model is unable to describe in full the mm-scale MRI signal generated based on an heterogenous and complex tissue micro-environment. The time-fractional order Bloch equations has been shown to provide, as a function of time, a good fit of brain MRI signals. We replaced the integer order Bloch equations with the previously reported time-fractional counterpart within the MRF framework and performed experiments to parcellate human gray matter, which is cortical brain tissue with different cyto-architecture at different spatial locations. Our findings suggest that the time-fractional order parameters, α and β, potentially associate with the effect of interareal architectonic variability, hypothetically leading to more accurate cortical parcellation.

physics.med-ph↗

arXiv2021

Embedding theory of lattices and its application for $2$-integrable lattices

For a positive integer $s$, a lattice $L$ is said to be $s$-integrable if $\sqrt{s}\cdot L$ is isometric to a sublattice of $\mathbb{Z}^n$ for some integer $n$. Conway and Sloane found two minimal non $2$-integrable lattices of rank $12$ and determinant $7$ in 1989. We find two more ones of rank $12$ and determinant $15$. Then we introduce a method of embedding a given lattice into a unimodular lattice, which plays a key role in proving minimality of non $2$-integrable lattices and finding candidates for non $2$-integrable lattices.

math.NT↗

arXiv2021

RCoNet: Deformable Mutual Information Maximization and High-order Uncertainty-aware Learning for Robust COVID-19 Detection

The novel 2019 Coronavirus (COVID-19) infection has spread world widely and is currently a major healthcare challenge around the world. Chest Computed Tomography (CT) and X-ray images have been well recognized to be two effective techniques for clinical COVID-19 disease diagnoses. Due to faster imaging time and considerably lower cost than CT, detecting COVID-19 in chest X-ray (CXR) images is preferred for efficient diagnosis, assessment and treatment. However, considering the similarity between COVID-19 and pneumonia, CXR samples with deep features distributed near category boundaries are easily misclassified by the hyper-planes learned from limited training data. Moreover, most existing approaches for COVID-19 detection focus on the accuracy of prediction and overlook the uncertainty estimation, which is particularly important when dealing with noisy datasets. To alleviate these concerns, we propose a novel deep network named {\em RCoNet$^k_s$} for robust COVID-19 detection which employs {\em Deformable Mutual Information Maximization} (DeIM), {\em Mixed High-order Moment Feature} (MHMF) and {\em Multi-expert Uncertainty-aware Learning} (MUL). With DeIM, the mutual information (MI) between input data and the corresponding latent representations can be well estimated and maximized to capture compact and disentangled representational characteristics. Meanwhile, MHMF can fully explore the benefits of using high-order statistics and extract discriminative features of complex distributions in medical imaging. Finally, MUL creates multiple parallel dropout networks for each CXR image to evaluate uncertainty and thus prevent performance degradation caused by the noise in the data.

eess.IV↗

arXiv2020

Recent progress on graphs with fixed smallest eigenvalue

We give a survey on graphs with fixed smallest eigenvalue, especially on graphs with large minimal valency and also on graphs with good structures. Our survey mainly consists of the following two parts: (i) Hoffman graphs, the basic theory related to Hoffman graphs and the applications of Hoffman graphs to graphs with fixed smallest eigenvalue and large minimal valency; (ii) recent results on distance-regular graphs and co-edge regular graphs with fixed smallest eigenvalue and the characterizations of certain families of distance-regular graphs. At the end of the survey, we also discuss signed graphs with fixed smallest eigenvalue and present some new findings.

math.CO↗

arXiv2020

Distributed Deep Convolutional Compression for Massive MIMO CSI Feedback

Massive multiple-input multiple-output (MIMO) systems require downlink channel state information (CSI) at the base station (BS) to achieve spatial diversity and multiplexing gains. In a frequency division duplex (FDD) multiuser massive MIMO network, each user needs to compress and feedback its downlink CSI to the BS. The CSI overhead scales with the number of antennas, users and subcarriers, and becomes a major bottleneck for the overall spectral efficiency. In this paper, we propose a deep learning (DL)-based CSI compression scheme, called DeepCMC, composed of convolutional layers followed by quantization and entropy coding blocks. In comparison with previous DL-based CSI reduction structures, DeepCMC proposes a novel fully-convolutional neural network (NN) architecture, with residual layers at the decoder, and incorporates quantization and entropy coding blocks into its design. DeepCMC is trained to minimize a weighted rate-distortion cost, which enables a trade-off between the CSI quality and its feedback overhead. Simulation results demonstrate that DeepCMC outperforms the state of the art CSI compression schemes in terms of the reconstruction quality of CSI for the same compression rate. We also propose a distributed version of DeepCMC for a multi-user MIMO scenario to encode and reconstruct the CSI from multiple users in a distributed manner. Distributed DeepCMC not only utilizes the inherent CSI structures of a single MIMO user for compression, but also benefits from the correlations among the channel matrices of nearby users to further improve the performance in comparison with DeepCMC. We also propose a reduced-complexity training method for distributed DeepCMC, allowing to scale it to multiple users, and suggest a cluster-based distributed DeepCMC approach for practical implementation.

eess.SP↗

arXiv2019

CNN-based Analog CSI Feedback in FDD MIMO-OFDM Systems

Massive multiple-input multiple-output (MIMO) systems require downlink channel state information (CSI) at the base station (BS) to better utilize the available spatial diversity and multiplexing gains. However, in a frequency division duplex (FDD) massive MIMO system, CSI feedback overhead degrades the overall spectral efficiency. Convolutional neural network (CNN)-based CSI feedback compression schemes has received a lot of attention recently due to significant improvements in compression efficiency; however, they still require reliable feedback links to convey the compressed CSI information to the BS. Instead, we propose here a CNN-based analog feedback scheme, called AnalogDeepCMC, which directly maps the downlink CSI to uplink channel input. Corresponding noisy channel outputs are used by another CNN to reconstruct the DL channel estimate. Not only the proposed outperforms existing digital CSI feedback schemes in terms of the achievable downlink rate, but also simplifies the operation as it does not require explicit quantization, coding and modulation, and provides a low-latency alternative particularly in rapidly changing MIMO channels, where the CSI needs to be estimated and fed back periodically.

cs.IT↗

arXiv2019

Mapping PM2.5 concentration at sub-km level resolution: a dual-scale retrieval method

Satellite-based retrieval has become a popular PM2.5 monitoring method currently. To improve the retrieval performance, multiple variables are usually introduced as auxiliary variable in addition to aerosol optical depth (AOD). Different kinds of variables are usually at different resolutions varying from sub-kilometers to dozens of kilometers. Generally, when doing the retrieval, variables at different resolutions are resampled to the same resolution as the AOD product to keep the scale consistency. A deficiency of doing this is that the information contained in the scale difference is discarded. To fully utilize the information contained at different scales, a dual-scale retrieval method is proposed in this study. At the first stage, variables which influence PM2.5 concentration at large scale were used for PM2.5 retrieval at coarse resolution. Then at the second stage, variables which affect PM2.5 distribution in finer scale, were used for the further PM2.5 retrieval at high resolution (sub-km level resolution) with the retrieved PM2.5 at the first stage at coarser resolution also as input. In this study, four different retrieval models including multiple linear regression (MLR), geographically weighted regression (GWR), random forest (RF) and generalized regression neural network (GRNN) are adopted to test the performance of the dual-scale retrieval method. Compared with the traditional retrieval method, the proposed dual-scale retrieval method can achieve PM2.5 mapping at finer resolution and with higher accuracy. Dual-scale retrieval can fully utilize the information contained at different scales, thus achieving a higher resolution and accuracy. It can be used for the generation of quantitative remote sensing products in various fields, and promote the improvement of the quality of quantitative remote sensing products.

physics.ao-ph↗