Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909555
Emilie D'Olne, Vincent W. Neo, P. Naylor
As the number of connected devices equipped with multiple microphones increases, scientific interest in distributed microphone array processing grows. Current beamforming methods heavily rely on estimating quantities related to array geometry, which is extremely challenging in real, non-stationary environments. Recent work on polynomial eigenvalue decomposition (PEVD) has shown promising results for speech enhancement in singular arrays without requiring the estimation of any array-related parameter [1]. This work extends these results to the realm of distributed microphone arrays, and further presents a novel framework for speech enhancement in distributed microphone arrays using PEVD. The proposed approach is shown to almost always outperform optimum beamformers located at arrays closest to the desired speaker. Moreover, the proposed approach exhibits very strong robustness to steering vector errors.
{"title":"Speech Enhancement in Distributed Microphone Arrays Using Polynomial Eigenvalue Decomposition","authors":"Emilie D'Olne, Vincent W. Neo, P. Naylor","doi":"10.23919/eusipco55093.2022.9909555","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909555","url":null,"abstract":"As the number of connected devices equipped with multiple microphones increases, scientific interest in distributed microphone array processing grows. Current beamforming methods heavily rely on estimating quantities related to array geometry, which is extremely challenging in real, non-stationary environments. Recent work on polynomial eigenvalue decomposition (PEVD) has shown promising results for speech enhancement in singular arrays without requiring the estimation of any array-related parameter [1]. This work extends these results to the realm of distributed microphone arrays, and further presents a novel framework for speech enhancement in distributed microphone arrays using PEVD. The proposed approach is shown to almost always outperform optimum beamformers located at arrays closest to the desired speaker. Moreover, the proposed approach exhibits very strong robustness to steering vector errors.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124141726","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909584
Xianrui Wang, J. Benesty, Gongping Huang, Jingdong Chen
Spectral estimation is of significant practical importance in a wide range of applications. This paper proposes a minimum variance distortionless response (MVDR) method for spectral estimation based on the Kronecker product. Taking advantage of the particular structure of the Fourier vector, we decompose it as a Kronecker product of two shorter vectors. Then, we design the spectral estimation filters under the same structure, i.e., as a Kronecker product of two filters. Consequently, the conventional MVDR spectrum problem is transformed to one of estimating two filters of much shorter lengths. Since it has much fewer parameters to estimate, the proposed method is able to achieve better performance than its conventional counterpart, particularly when the number of available signal samples is small. Also presented in this paper is the generalization to the estimation of the cross-spectrum and coherence function.
{"title":"A Minimum Variance Distortionless Response Spectral Estimator with Kronecker Product Filters","authors":"Xianrui Wang, J. Benesty, Gongping Huang, Jingdong Chen","doi":"10.23919/eusipco55093.2022.9909584","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909584","url":null,"abstract":"Spectral estimation is of significant practical importance in a wide range of applications. This paper proposes a minimum variance distortionless response (MVDR) method for spectral estimation based on the Kronecker product. Taking advantage of the particular structure of the Fourier vector, we decompose it as a Kronecker product of two shorter vectors. Then, we design the spectral estimation filters under the same structure, i.e., as a Kronecker product of two filters. Consequently, the conventional MVDR spectrum problem is transformed to one of estimating two filters of much shorter lengths. Since it has much fewer parameters to estimate, the proposed method is able to achieve better performance than its conventional counterpart, particularly when the number of available signal samples is small. Also presented in this paper is the generalization to the estimation of the cross-spectrum and coherence function.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122629216","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909610
Marcelo Fiori, Bernardo Marenco, Federico Larroca, P. Bermolen, G. Mateos
The Random Dot Product Graph (RDPG) is a popular generative graph model for relational data. RDPGs postulate there exist latent positions for each node, and specifies the edge formation probabilities via the inner product of the corresponding latent vectors. The embedding task of estimating these latent positions from observed graphs is usually posed as a non-convex matrix factorization problem. The workhorse Adjacency Spectral Embedding offers an approximate solution obtained via the eigendecomposition of the adjacency matrix, which enjoys solid statistical guarantees but can be computationally intensive and is formally solving a surrogate problem. In this paper, we bring to bear recent non-convex optimization advances and demonstrate their impact to RDPG inference. We develop first-order gradient descent methods to better solve the original optimization problem, and to accommodate broader network embedding applications in an organic way. The effectiveness of the resulting graph representation learning framework is demonstrated on both synthetic and real data. We show the algorithms are scalable, robust to missing network data, and can track the latent positions over time when the graphs are acquired in a streaming fashion.
{"title":"Algorithmic Advances for the Adjacency Spectral Embedding","authors":"Marcelo Fiori, Bernardo Marenco, Federico Larroca, P. Bermolen, G. Mateos","doi":"10.23919/eusipco55093.2022.9909610","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909610","url":null,"abstract":"The Random Dot Product Graph (RDPG) is a popular generative graph model for relational data. RDPGs postulate there exist latent positions for each node, and specifies the edge formation probabilities via the inner product of the corresponding latent vectors. The embedding task of estimating these latent positions from observed graphs is usually posed as a non-convex matrix factorization problem. The workhorse Adjacency Spectral Embedding offers an approximate solution obtained via the eigendecomposition of the adjacency matrix, which enjoys solid statistical guarantees but can be computationally intensive and is formally solving a surrogate problem. In this paper, we bring to bear recent non-convex optimization advances and demonstrate their impact to RDPG inference. We develop first-order gradient descent methods to better solve the original optimization problem, and to accommodate broader network embedding applications in an organic way. The effectiveness of the resulting graph representation learning framework is demonstrated on both synthetic and real data. We show the algorithms are scalable, robust to missing network data, and can track the latent positions over time when the graphs are acquired in a streaming fashion.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127981128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909830
Alex Cazañas-Gordón, Luís A. da Silva Cruz
This study presents a novel approach to detecting the retina in optical coherence tomography (OCT) images using Deep Q learning. The proposed method uses an agent to extract contextual information from the input OCT to produce a tight-bounding box around the retina in a step-wise fashion. The detection task implements a decision process governed by a reinforcement learning strategy, where the agent takes actions and receives rewards according to their outcome. During the localization process, the agent learns the optimal set of actions to complete the detection task using a Q-network that estimates the value of the expected return of an action at any given step. Experiments on a test OCT dataset of 100 images showed that the proposed method accurately located the retina with a mean recall of 0.988 and a mean F1 score of 0.94.
{"title":"Automatic Detection of the Retina in Optical Coherence Tomography using Deep Q Learning","authors":"Alex Cazañas-Gordón, Luís A. da Silva Cruz","doi":"10.23919/eusipco55093.2022.9909830","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909830","url":null,"abstract":"This study presents a novel approach to detecting the retina in optical coherence tomography (OCT) images using Deep Q learning. The proposed method uses an agent to extract contextual information from the input OCT to produce a tight-bounding box around the retina in a step-wise fashion. The detection task implements a decision process governed by a reinforcement learning strategy, where the agent takes actions and receives rewards according to their outcome. During the localization process, the agent learns the optimal set of actions to complete the detection task using a Q-network that estimates the value of the expected return of an action at any given step. Experiments on a test OCT dataset of 100 images showed that the proposed method accurately located the retina with a mean recall of 0.988 and a mean F1 score of 0.94.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128101260","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909969
Xiaohan Zhao, Yongzhe Li, R. Tao
In this paper, we focus on the unimodular waveform design with good correlation property, i.e., with low integrated sidelobe level (ISL). In contrast to existing approaches that commonly involve constraints on the moduli of waveform elements, we come up with the idea of designing the waveform via directly optimizing its phase values. Using this idea, the standard ISL-minimization based waveform design is converted as an unconstrained optimization problem with respect to the phase values of waveform elements, which avoids the repetitive procedure of projecting non-unimodular complex values into the best approximations of constant magnitudes. To this end, we first reformulate the ISL metric into a function of the phase values to be obtained for the waveform, and then solve the new unconstrained ISL-minimization-based waveform design using majorization-minimization techniques. The first-order gradient of the reformulated objective function is derived, by which the majorant of the objective is elaborated. Based on this, we finally tackle the design via iterations, at each of which we obtain a closed-form solution with fast implementations. An algorithm is proposed, with whose simpleness and effectiveness are verified by simulations.
{"title":"Design of Single Unimodular Waveform With Good Correlation Level Via Phase Optimizations","authors":"Xiaohan Zhao, Yongzhe Li, R. Tao","doi":"10.23919/eusipco55093.2022.9909969","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909969","url":null,"abstract":"In this paper, we focus on the unimodular waveform design with good correlation property, i.e., with low integrated sidelobe level (ISL). In contrast to existing approaches that commonly involve constraints on the moduli of waveform elements, we come up with the idea of designing the waveform via directly optimizing its phase values. Using this idea, the standard ISL-minimization based waveform design is converted as an unconstrained optimization problem with respect to the phase values of waveform elements, which avoids the repetitive procedure of projecting non-unimodular complex values into the best approximations of constant magnitudes. To this end, we first reformulate the ISL metric into a function of the phase values to be obtained for the waveform, and then solve the new unconstrained ISL-minimization-based waveform design using majorization-minimization techniques. The first-order gradient of the reformulated objective function is derived, by which the majorant of the objective is elaborated. Based on this, we finally tackle the design via iterations, at each of which we obtain a closed-form solution with fast implementations. An algorithm is proposed, with whose simpleness and effectiveness are verified by simulations.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129848310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909725
Esther Rodrigo Bonet, T. Do, Xuening Qin, J. Hofman, V. Manna, Wilfried Philips, Nikos Deligiannis
To control air pollution and mitigate its negative effect on health, it is of the utmost importance to have accurate real-time forecasting models. Existing deep-learning-based air quality forecasting models typically deploy temporal and-less often-spatial modules. Yet, data scarcity emerges as a real issue in this domain, a problem that can be solved by capturing the data distribution. In this work, we address data scarcity by proposing a novel conditional variational graph autoencoder. Our model is able to forecast air pollution by efficiently encoding the spatio-temporal correlations of the known data. Additionally, we leverage dynamic context data such as weather or satellite images to condition the model's behaviour. We formulate the problem as a context-aware graph-based matrix completion task and utilize street-level data from mobile stations. Experiments on real-world air quality datasets show the improved performance of our model with respect to state-of-the-art approaches.
{"title":"Conditional Variational Graph Autoencoder for Air Quality Forecasting","authors":"Esther Rodrigo Bonet, T. Do, Xuening Qin, J. Hofman, V. Manna, Wilfried Philips, Nikos Deligiannis","doi":"10.23919/eusipco55093.2022.9909725","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909725","url":null,"abstract":"To control air pollution and mitigate its negative effect on health, it is of the utmost importance to have accurate real-time forecasting models. Existing deep-learning-based air quality forecasting models typically deploy temporal and-less often-spatial modules. Yet, data scarcity emerges as a real issue in this domain, a problem that can be solved by capturing the data distribution. In this work, we address data scarcity by proposing a novel conditional variational graph autoencoder. Our model is able to forecast air pollution by efficiently encoding the spatio-temporal correlations of the known data. Additionally, we leverage dynamic context data such as weather or satellite images to condition the model's behaviour. We formulate the problem as a context-aware graph-based matrix completion task and utilize street-level data from mobile stations. Experiments on real-world air quality datasets show the improved performance of our model with respect to state-of-the-art approaches.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129973693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909715
Christoph Brauer, D. Lorenz, Lionel Tondji
The incorporation of prior knowledge into the ma-chine learning pipeline is subject of informed machine learning. Spatial invariances constitute a class of prior knowledge that can be taken into account especially in the design of model architectures or through virtual training examples. In this contribution, we investigate fully connected neural network architectures that are equivariant with respect to the dihedral group of order eight. This is practically motivated by the application of leakage detection in vacuum bagging which plays an important role in the manufacturing of fiber composite components. Our approach for the derivation of an equivariant architecture is constructive and transferable to other symmetry groups. It starts from a standard network architecture and results in a specific kind of weight sharing in each layer. In numerical experiments, we compare equivariant and standard networks on a novel leakage detection dataset. Our results indicate that group equivariant networks can capture the application specific prior knowledge much better than standard networks, even if the latter are trained on augmented data.
{"title":"Group equivariant networks for leakage detection in vacuum bagging","authors":"Christoph Brauer, D. Lorenz, Lionel Tondji","doi":"10.23919/eusipco55093.2022.9909715","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909715","url":null,"abstract":"The incorporation of prior knowledge into the ma-chine learning pipeline is subject of informed machine learning. Spatial invariances constitute a class of prior knowledge that can be taken into account especially in the design of model architectures or through virtual training examples. In this contribution, we investigate fully connected neural network architectures that are equivariant with respect to the dihedral group of order eight. This is practically motivated by the application of leakage detection in vacuum bagging which plays an important role in the manufacturing of fiber composite components. Our approach for the derivation of an equivariant architecture is constructive and transferable to other symmetry groups. It starts from a standard network architecture and results in a specific kind of weight sharing in each layer. In numerical experiments, we compare equivariant and standard networks on a novel leakage detection dataset. Our results indicate that group equivariant networks can capture the application specific prior knowledge much better than standard networks, even if the latter are trained on augmented data.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130457233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909702
Thanh Trung LE, K. Abed-Meraim, N. Trung, A. Hafiane
Robust tensor tracking or robust adaptive tensor decomposition of streaming tensors is crucial when observations are corrupted by sparse outliers and missing data. In this paper, we introduce a novel tensor tracking algorithm for factorizing incomplete streaming tensors with sparse outliers under tensor-train (TT) format. The proposed algorithm consists of two main stages: online outlier rejection and tracking of TT-cores. In the former stage, outliers affecting the data streams are efficiently detected by an ADMM solver. In the latter stage, we propose an effective recursive least-squares solver to incrementally update TT-cores at each time $t$. Several numerical experiments on both simulated and real data are presented to verify the effectiveness of the proposed algorithm.
{"title":"Robust Tensor Tracking With Missing Data Under Tensor-Train Format","authors":"Thanh Trung LE, K. Abed-Meraim, N. Trung, A. Hafiane","doi":"10.23919/eusipco55093.2022.9909702","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909702","url":null,"abstract":"Robust tensor tracking or robust adaptive tensor decomposition of streaming tensors is crucial when observations are corrupted by sparse outliers and missing data. In this paper, we introduce a novel tensor tracking algorithm for factorizing incomplete streaming tensors with sparse outliers under tensor-train (TT) format. The proposed algorithm consists of two main stages: online outlier rejection and tracking of TT-cores. In the former stage, outliers affecting the data streams are efficiently detected by an ADMM solver. In the latter stage, we propose an effective recursive least-squares solver to incrementally update TT-cores at each time $t$. Several numerical experiments on both simulated and real data are presented to verify the effectiveness of the proposed algorithm.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129751898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909654
Orkény Zováthi, Balázs Pálffy, C. Benedek
In this paper we introduce a novel approach for accurate self-localization and pose tracking for Lidar and GPS-equipped autonomous vehicles (AVs) in high-density (more than 5000 points/m2) 3D localization maps obtained through Mobile Laser Scanning (MLS). Our solution consist of two main steps: First, starting from a poor GPS-based initial position, we estimate the 3DoF pose (planar position and yaw orientation) of the ego vehicle by aligning its sparse (50-500 points/m2) Lidar point cloud measurements to the MLS prior map, using a novel approach of matching static landmark objects of the scene. Second, to effectively deal with the lack of pairable objects in certain time frames (e.g. due to scene segments occluded by a large moving tram), we track the estimated 3DoF pose of the AVs by a Kalman filter. Comperative test are provided on roads with heavy traffic in downtown city areas with large (5-10 meters) GPS positioning errors. The proposed approach is able to reduce the location error of the vehicle by one order of magnitude and keep the yaw angle error around 1° during its whole trajectory without considerable drift, while running in real-time (20-25 Hz).
{"title":"Real-time Vehicle Localization and Pose Tracking in High-Resolution 3D Maps","authors":"Orkény Zováthi, Balázs Pálffy, C. Benedek","doi":"10.23919/eusipco55093.2022.9909654","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909654","url":null,"abstract":"In this paper we introduce a novel approach for accurate self-localization and pose tracking for Lidar and GPS-equipped autonomous vehicles (AVs) in high-density (more than 5000 points/m2) 3D localization maps obtained through Mobile Laser Scanning (MLS). Our solution consist of two main steps: First, starting from a poor GPS-based initial position, we estimate the 3DoF pose (planar position and yaw orientation) of the ego vehicle by aligning its sparse (50-500 points/m2) Lidar point cloud measurements to the MLS prior map, using a novel approach of matching static landmark objects of the scene. Second, to effectively deal with the lack of pairable objects in certain time frames (e.g. due to scene segments occluded by a large moving tram), we track the estimated 3DoF pose of the AVs by a Kalman filter. Comperative test are provided on roads with heavy traffic in downtown city areas with large (5-10 meters) GPS positioning errors. The proposed approach is able to reduce the location error of the vehicle by one order of magnitude and keep the yaw angle error around 1° during its whole trajectory without considerable drift, while running in real-time (20-25 Hz).","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131089240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-29DOI: 10.23919/eusipco55093.2022.9909566
Yang Li, G. Mateos
Advances in graph signal processing for network neuroscience offer a unique pathway to integrate brain structure and function, with the goal of revealing some of the brain's organizing principles at the system level. In this direction, we develop a supervised graph representation learning framework to model the relationship between brain structural connectivity (SC) and functional connectivity (FC) via a graph encoder-decoder system. Specifically, we propose a Siamese network architecture equipped with graph convolutional encoders to learn graph (i.e., subject)-level embeddings that preserve application-dependent similarity measures between brain networks. This way, we effectively increase the number of training samples and bring in the flexibility to incorporate additional prior information via the prescribed target graph-level distance. While information on the brain structure-function coupling is implicitly distilled via reconstruction of brain FC from SC, our model also manages to learn representations that preserve the similarity between input graphs. The superior discriminative power of the learnt representations is demonstrated in downstream tasks including subject classification and visualization. All in all, this work advocates the prospect of leveraging learnt graph-level, similarity-preserving embeddings for brain network analysis, by bringing to bear standard tools of metric data analysis.
{"title":"Learning Similarity-Preserving Representations of Brain Structure-Function Coupling","authors":"Yang Li, G. Mateos","doi":"10.23919/eusipco55093.2022.9909566","DOIUrl":"https://doi.org/10.23919/eusipco55093.2022.9909566","url":null,"abstract":"Advances in graph signal processing for network neuroscience offer a unique pathway to integrate brain structure and function, with the goal of revealing some of the brain's organizing principles at the system level. In this direction, we develop a supervised graph representation learning framework to model the relationship between brain structural connectivity (SC) and functional connectivity (FC) via a graph encoder-decoder system. Specifically, we propose a Siamese network architecture equipped with graph convolutional encoders to learn graph (i.e., subject)-level embeddings that preserve application-dependent similarity measures between brain networks. This way, we effectively increase the number of training samples and bring in the flexibility to incorporate additional prior information via the prescribed target graph-level distance. While information on the brain structure-function coupling is implicitly distilled via reconstruction of brain FC from SC, our model also manages to learn representations that preserve the similarity between input graphs. The superior discriminative power of the learnt representations is demonstrated in downstream tasks including subject classification and visualization. All in all, this work advocates the prospect of leveraging learnt graph-level, similarity-preserving embeddings for brain network analysis, by bringing to bear standard tools of metric data analysis.","PeriodicalId":231263,"journal":{"name":"2022 30th European Signal Processing Conference (EUSIPCO)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125873113","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：