How Long short-term memory artificial neural network, synthetic data, and fine-tuning improve the classification of raw EEG data

Albert Nasybullin; Vladimir Maksimenko; Semen Kurkin

cs.LGApr 5, 2026

How Long short-term memory artificial neural network, synthetic data, and fine-tuning improve the classification of raw EEG data

Authors: Albert Nasybullin, Vladimir Maksimenko, Semen Kurkin

Organizations: Laboratory of Data Analysis and Bioinformatics, Center for Artificial Intelligence, Innopolis University, Innopolis, Russia · Neuroscience and Cognitive Technology Laboratory, Center for Technologies in Robotics and Mechatronics Components, Innopolis University, Innopolis, Russia

Abstract

In this paper, we discuss a Machine Learning pipeline for the classification of EEG data. We propose a combination of synthetic data generation, long short-term memory artificial neural network (LSTM), and fine-tuning to solve classification problems for experiments with implicit visual stimuli, such as the Necker cube with different levels of ambiguity. The developed approach increased the quality of the classification model of raw EEG data.

Figures & tables

Fig. 1: Ambiguous values of the Necker cube in visual stimulus [ 9 ]

Class labels	”Left”	”High-ambiguity”	”Right”
Ambiguity values	0.15, 0.25	0.4, 0.45, 0.55, 0.6	0.75, 0.85
Total dataset size	997	2 000	1 003
Train subset size	598	1 200	602
Test subset size	399	800	401

TABLE I: Original dataset structure

Fig. 2: The architecture of the Long short-term memory artificial neural network model used for classification.

Fig. 3: We build our intuition behind data expanding strategy on a difference created in a signal by band-pass filtering.

	Training data subset
F-1 score	Original	Theta	Alpha	Beta
0.63	+
0.78	+	+
0.71	+		+
0.69	+			+
0.73	+	+	+	+

TABLE II: Model evaluation with different pre-trained weights

EEG Decoding

Electroencephalography (EEG) decoding maps brain-signal patterns to mental states, stimuli, speech, or intended actions.

See all 208 papers

Generative Augmentation for EEG Motor Imagery Classification: A Class-Conditional VAE with Cycle-Consistent Decoder Refinement
Jul 22, 2026Matei Moldoveanu, Alain Sirois, Claire Ben Ali +2Synthetic Data Augmentation MI Classification
Naqi Logix, Paris, France · Inria Center at Univ. Bordeaux / LaBRI, Talence, France · LITIS, INSA Rouen-Normandy, France
Convolutional Neural Network and Adversarial Autoencoder in EEG images classification
Apr 5, 2026Albert Nasybullin, Semen KurkinEEG Decoding MI Classification
Faculty of Computer Science and Engineering, Innopolis University, Innopolis, Russia · Neuroscience and Cognitive Technology Laboratory, Center for Technologies in Robotics and Mechatronics Components, Innopolis University, Innopolis, Russia
GraM-Diff: A Unified Graph-Mamba Diffusion Framework for EEG-Based Alzheimer's Disease Data Generation and Diagnosis
Aug 30, 2026M. Tanveer, Ayush Singh Rana, Sanskriti Jain +5Synthetic Data Augmentation Electroencephalography
Indian Institute of Technology Indore, Indore 453552, Madhya Pradesh, India

Jul 22, 2026J/K move · Enter open · S save

Generative Augmentation for EEG Motor Imagery Classification: A Class-Conditional VAE with Cycle-Consistent Decoder Refinement

Matei Moldoveanu, Alain Sirois, Claire Ben Ali, Fabien Lotte+1

Naqi Logix, Paris, France · Inria Center at Univ. Bordeaux / LaBRI, Talence, France · LITIS, INSA Rouen-Normandy, France

Synthetic Data Augmentation MI Classification Conditional Variational Autoencoder EEG Signal Generation

Open paper AskPDF arXiv

We investigate whether a generative model can supply useful synthetic motor-imagery (MI) electroencephalography (EEG) trials that improve the accuracy of independent downstream classifiers. We train a class-conditional variational autoencoder (CVAE) with an integrated latent classifier on the Zhou motor-imagery dataset, using the learned per-class prior as a generator: sampling the prior for a given label and decoding it into a synthetic, label-consistent signal. A constraint on the covariance matrix of the generated data encourages preservation of covariance structure, and the model is trained with a schedule that alternates ordinary VAE training with a decoder-focused phase that sharpens the generative pathway used for augmentation. We measure the effect of adding synthetic trials to the training set under two evaluation protocols -- within-user (pooled 60/20/20 split across subjects) and cross-user (leave-one-subject-out, LOSO) -- across four representative EEG classification pipelines: Common Spatial Patterns with Linear Discriminant Analysis (CSP+LDA), tangent-space features with a Support Vector Machine (TGSP+SVM), Minimum Distance to Riemannian Mean (MDM), and a neural network based on EEGNetv4 (henceforth EEGNet). Results are aggregated across independent augmentation draws, random seeds (within-user), or leave-one-subject-out folds (cross-user), with uncertainty reported as 95% confidence intervals (Student's

t

-distribution) computed over per-seed/per-fold averages. We find that synthetic EEG from the CVAE is most credible as a source of class-structured, covariance-like data rather than as a substitute for real raw EEG: it can raise the point estimate for MDM, but the broader augmentation claim remains conservative -- observed gains are small and classifier-dependent.

Abstract

EEG Decoding

Explore similar work

Generative Augmentation for EEG Motor Imagery Classification: A Class-Conditional VAE with Cycle-Consistent Decoder Refinement

Convolutional Neural Network and Adversarial Autoencoder in EEG images classification

GraM-Diff: A Unified Graph-Mamba Diffusion Framework for EEG-Based Alzheimer's Disease Data Generation and Diagnosis