Structure-Preserving Transformers for Sequences of SPD Matrices

被引：0

作者：

Seraphim, Mathieu ^{[1
]}

Lechervy, Alexis ^{[1
]}

Yger, Florian ^{[2
]}

Brun, Luc ^{[1
]}

Etard, Olivier ^{[3
]}

机构：

[1] Normandie Univ, UNICAEN, ENSICAEN, CNRS,GREYC, F-14000 Caen, France

[2] PSL Univ Paris Dauphine, CNRS, LAMSADE, F-75016 Paris, France

[3] Normandie Univ, UNICAEN, CHU Caen, INSERM,COMETE,CYCERON, F-14000 Caen, France

来源：

32ND EUROPEAN SIGNAL PROCESSING CONFERENCE, EUSIPCO 2024 | 2024年

关键词：

Transformers; SPD Matrices; Structure-Preserving; Electroencephalography; Sleep Staging; MANIFOLD; SLEEP;

D O I：

10.23919/EUSIPCO63174.2024.10715089

中图分类号：

TP31 [计算机软件];

学科分类号：

081202 ; 0835 ;

摘要：

In recent years, Transformer-based auto-attention mechanisms have been successfully applied to the analysis of a variety of context-reliant data types, from texts to images and beyond, including data from non-Euclidean geometries. In this paper, we present such a mechanism, designed to classify sequences of Symmetric Positive Definite matrices while preserving their Riemannian geometry throughout the analysis. We apply our method to automatic sleep staging on timeseries of EEG-derived covariance matrices from a standard dataset, obtaining high levels of stage-wise performance.

引用

页码：1451 / 1455

页数：5

共 29 条

[1] Optuna: A Next-generation Hyperparameter Optimization Framework [J].

Akiba, Takuya ;

Sano, Shotaro ;

Yanase, Toshihiko ;

Ohta, Takeru ;

Koyama, Masanori .

KDD'19: PROCEEDINGS OF THE 25TH ACM SIGKDD INTERNATIONAL CONFERENCCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 2019, :2623-2631

[2] Log-euclidean metrics for fast and simple calculus on diffusion tensors [J].

Arsigny, Vincent ;

Fillard, Pierre ;

Pennec, Xavier ;

Ayache, Nicholas .

MAGNETIC RESONANCE IN MEDICINE, 2006, 56 (02) :411-421

[3] AASM Scoring Manual Updates for 2017 (Version 2.4) [J].

Berry, Richard B. ;

Brooks, Rita ;

Gamaldo, Charlene ;

Harding, Susan M. ;

Lloyd, Robin M. ;

Quan, Stuart F. ;

Troester, Matthew T. ;

Vaughn, Bradley V. .

JOURNAL OF CLINICAL SLEEP MEDICINE, 2017, 13 (05) :665-666

[4] ManifoldNet: A Deep Neural Network for Manifold-Valued Data With Applications [J].

Chakraborty, Rudrasis ;

Bouza, Jose ;

Manton, Jonathan ;

Vemuri, Baba C. .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (02) :799-810

[5] Automatic Sleep Stage Classification on EEG Signals Using Time-Frequency Representation [J].

Dequidt, Paul ;

Seraphim, Mathieu ;

Lechervy, Alexis ;

Gaez, Ivan Igor ;

Brun, Luc ;

Etard, Olivier .

ARTIFICIAL INTELLIGENCE IN MEDICINE, AIME 2023, 2023, 13897 :250-259

[6]

Eickhoff SB., 2015, BRAIN MAPPING, P187, DOI DOI 10.1016/B978-0-12-397025-1.00212-8

[7] RobustSleepNet: Transfer Learning for Automated Sleep Staging at Scale [J].

Guillot, Antoine ;

Thorey, Valentin .

IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, 2021, 29 :1441-1451

[8] Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (CMA-ES) [J].

Hansen, N ;

Muller, SD ;

Koumoutsakos, P .

EVOLUTIONARY COMPUTATION, 2003, 11 (01) :1-18

[9]

Harandi MT, 2014, LECT NOTES COMPUT SC, V8690, P17, DOI 10.1007/978-3-319-10605-2_2

[10]

He L, 2021, Advances in Neural Information Processing Systems, P343

← 1 2 3 →