Awesome-AI4BCI

AI, especially Deep Learning, has made breakthroughs in learning from Brain Signals, vital for both Brain Encoding and Decoding. Unlock the potential with this repository—a curated collection of resources and papers on Deep Learning Models for Brain-Computer Interfaces. Dive in to explore the future of brain-related AI advancements.

Contents

• Papers
  • Survey
  • FSL for BCIs
  • MetaL for BCI
  • Equivariant Networks
  • Transformer for BCI
  • Materials Generation
  • Awesome Githubs
  • Resources-for-Students
  • Architectures

Papers

New arrivals

• Brant-2: Foundation Model for Brain Signals

###General

• Towards Zero Training for Brain-Computer Interfacing
• Evaluation Criteria for BCI Research

Survey

• International Winter Workshop on Brain-Computer Interface (BCI)]
  _{IEEE Xplore & 2023.}
• 2020 International brain–computer interface competition: A review]
  _{Ji-Hoon Jeong et.al china; Frontiers of Human Neuroscience & 2022.}

Architectures

• Survey

• On the Deep Learning Models for EEG-Based Brain-Computer Interface Using Motor Imagery
• MAtt: A Manifold Attention Network for EEG Decoding - geometric learning
• EEG-ITNet: An Explainable Inception Temporal Convolutional Network for Motor Imagery Classification
• MIN2Net: End-to-End Multi-Task Learning for Subject-Independent Motor Imagery EEG Classification - Supple - (Tool Python API)

Multi-task leanring for BCI

• Multitask Learning for Brain-Computer Interfaces
• Multi-Task Logistic Regression in Brain-Computer Interfaces
• Transfer Learning in Brain-Computer Interfaces

Transfer-Learning for EEG

• Adaptive transfer learning for EEG motor imagery classification with deep Convolutional Neural Network
• EEG based multi-class seizure type classification using convolutional neural network and transfer learning
• Transfer learning promotes acquisition of individual BCI skills

FSL4BCI

MI-BCI

• Dual Attention Relation Network With Fine-Tuning for Few-Shot EEG Motor Imagery Classification
• Decoding Multi-Brain Motor Imagery from EEG Using Coupling Feature Extraction and Few-Shot Learning
  _{Li Zhu, Youyang Liu, Riheng Liu, Yong Peng, Jianting Cao,Junhua Li, and Wanzeng Kong, Senior Member, IEEE* IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING & 23/11/2023.}
• A Few-Shot Transfer Learning Approach for Motion Intention Decoding from Electroencephalographic Signals
• Decoding Multi-Brain Motor Imagery From EEG Using Coupling Feature Extraction and Few-Shot Learning
• Meta-Optimization of Initial Weights for More Effective Few- and Zero-Shot Learning in BCI Classification
• Zero-Shot Learning for EEG Classification in Motor Imagery-Based BCI System
• Cross-Subject EEG Emotion Recognition Using Domain Adaptive Few-Shot Learning Networks
• Decoding Multi-Brain Motor Imagery From EEG Using Coupling Feature Extraction and Few-Shot Learning
• FSL:2020 International brain–computer interface competition: A review

non MI-BCI

• EEG-Fest: Few-shot based Attention Network for Driver's DrowsinessEstimation with EEG Signals
  _{Ning Ding, Ce Zhang and Azim Eskandarian, Mechanical Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA -Biomedical Physics & Engineering 28/11/2023}
• A prototypical network for few-shot recognition of speech imagery data
• Prediction of Preoperative Scale Score of Dystonia Based on Few-Shot Learning
• Emotion Recognition from Few-Channel EEG Signals by Integrating Deep Feature Aggregation and Transfer Learning]
  _{*Fang Liu et.al, BNRist, the Department of Computer Science and Technology, Tsinghua University, China
  IEEE Transactions on Affective Computing & 21/11/2023.}
• Neonatal seizure detection combined deep network and meta-learning

MetaLearning4BCI

Bayesian MetaL

• Adaptive Bayesian Meta-Learning for EEG Signal Classification

MetaL for MI-BCI

• META-EEG: Meta-learning-based class-relevant EEG representation learning for zero-calibration brain–computer interfaces]
  _{Ji-Wung Han1, Soyeon Bak, Jun-Mo Kim, WooHyeok Choi, Dong-Hee Shin, Young-Han Son,Tae-Eui Kam ∗ Department of Artificial Intelligence, Korea University, Seoul 02841, South Korea; Expert Systems With Applications & 10/10/2023.}
• Does Meta-Learning Improve EEG Motor Imagery Classification
  _{Xiaoli Wu, and Rosa H. M. Chan, Senior Member EMBC & 2022.}
• Meta Learn on Constrained Transfer Learning for Low Resource Cross Subject EEG Classification
  _{TIEHANG DUAN 1,2, MOHAMMAD ABUZAR SHAIKH1, MIHIR CHAUHAN1, JUN CHU1,ROHINI K. SRIHARI, ARCHITA PATHAK1, AND SARGUR N. SRIHARI -IEEE Access & 16/12/2020.}
• Ultra Efficient Transfer Learning with Meta Update for Continuous EEG Classification Across Subjects*
  _{Tiehang Duan, Mihir Chauhan, Mohammad Abuzar Shaikh, Jun Chu, and Sargur N. Srihari journal & 2018}
• Meta-Learning for Fast and Privacy-Preserving Source Knowledge Transfer of EEG-Based BCIs
• Privacy-Preserving Domain Adaptation for Motor Imagery-Based Brain-Computer Interfaces
• On the Vulnerability of CNN Classifiers in EEG-Based BCIs
• Subject-independent meta-learning framework towards optimal training of EEG-based classifiers
• Model Agnostic Meta Learning for EEG Classification: Multitask Approach
• Few Is Enough: Task-Augmented Active Meta-Learning for Brain Cell Classification(https://arxiv.org/pdf/2007.05009.pdf)
• Convolution Monge Mapping Normalization for learning on sleep data

MetaL4 non MI-EEG BCI

• Calibration free meta learning based approach for subject independent EEG emotion recognition
  _{Swapnil Bhosale, Rupayan Chakraborty, Sunil Kumar Kopparapu, TCS Research | Biomedical Signal Processing and Control, 2022}
• MetaEmotionNet: Spatial–Spectral–Temporal-Based Attention 3-D Dense Network With Meta-Learning for EEG Emotion Recognition

Domain Adaptation in EEG

• Learning a robust unified domain adaptation framework for cross-subject EEG-based emotion recognition]
  _{Magdiel Jiménez-Guarneros , Gibran Fuentes-Pineda -Biomedical Signal Processing and Control & 22/06/2023.}

Transformer4BCI

• paper

ExAI4BCI

• Interpretable deep neural networks for single-trial EEG classification
  _{Irene Sturma, Sebastian Lapuschkinb, Wojciech Samek b,∗, Klaus-Robert Müller a,c, a Machine Learning Group, Berlin - Journal of Neuroscience Methods & 16/10/2016.}

• XBrainLab: An Open-Source Software for Explainable Artificial Intelligence-Based EEG Analysis -
  _{introduce XBrainLab, an open-source user-friendly software, for accelerated interpretation of neural patterns from EEG data based on cutting-edge computational approach. More details in this [[presentation] (https://docs.google.com/presentation/d/17N0tN0pdMqVCeIFpMDHl6H4B92INjt-OYddn6me4Nds/edit#slide=id.g2423a157453_1_0)]}

Addressing Problems of BCI

Variability in EEG-BCI

• Discrepancy between inter- and intra-subject variability in EEG-based motor imagery brain-computer interface: Evidence from multiple perspectives]
  _{Gan Huang,Zhiheng Zhao1,Shaorong Zhang,Zhenxing Hu1,Jiaming Fan1,Meisong Fu1,Jiale Chen1,Yaqiong Xiao4, Jun Wang5 and Guo Dan; China - Frontiers in Neuroscience & 13/02/2023.}
• Reducing the Subject Variability of EEG Signals with Adversarial Domain Generalization]
  _{Bo-Qun Ma1, He Li1, Wei-Long Zheng2, and Bao-Liang Lu1, China - Springer Nature & 2019.}
• MITIGATING INTER-SUBJECT BRAIN SIGNAL VARIABILITY FOR EEG-BASED DRIVER FATIGUE STATE CLASSIFICATION]
  _{Sunhee Hwang, Sungho Park, Dohyung Kim, Jewook Lee1 and Hyeran Byun - ICASSP & 2021}
• [EEG variability: Task-driven or subject-driven signal of interest?]https://www.sciencedirect.com/science/article/pii/S105381192200163X)]
  _{Erin Gibson a, Nancy J. Lobaugh c d, Steve Joordens b, Anthony R. McIntosh,Univ of Totanto Canada; Elsevier & 15 MAy 2022}
• Inter-subject transfer learning with an end-to-end deep convolutional neural network for EEG-based BC] - Journal of Neural Engineering & 2019.
• paper

Zero-Calibration BCI

• Switching EEG Headsets Made Easy: Reducing Offline Calibration Effort Using Active Weighted Adaptation Regularization** ]
  _{Dongrui Wu, Vernon J. Lawhern, W. David Hairston, and Brent J. Lance, Senior Member, IEEE - IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, & 2016.}

Generalised Neural Decoder

• Generalized neural decoders for transfer learning across participants and recording modalities]
  _{Steven M Peterson1,2, Zoe Steine-Hanson3, Nathan Davis, Rajesh P N Rao,and Bingni W Brunton | journal of Neural Engineering & 2020.}

BCI in top AI conferences and journals (NeurIPS, ICML, ICLR, ACL, EMNLP, ICCV, CVPR, TMLR, JMLR, TACL, AAAI)

• NeurIPS2024 | CSLP-AE: A Contrastive Split-Latent Permutation Autoencoder Framework for Zero-Shot Electroencephalography Signal Conversion
• [Generalizable Movement Intention Recognition with Multiple Heterogeneous EEG Datasets(]https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10160462&tag=1)
• EEG Decoding for Datasets with Heterogenous Electrode Configurations using Transfer Learning Graph Neural Networks
• Learning Topology-Agnostic EEG Representations with Geometry-Aware Modeling
• ICLR2024 LARGE BRAIN MODEL FOR LEARNING GENERIC REPRESENTATIONS WITH TREMENDOUS EEG DATA IN BCI(LLM)

DataSet

• A large EEG dataset for studying cross-session variability in motor imagery brain-computer interface]
  _{Jun Ma, Banghua Yang, Wenzheng Qiu, Yunzhe Li, Shouwei Gao & Xinxing Xia Nature & 01/09/2022.}
• BCI Competition 2008 – Graz data set A]
  _{C. Brunner, R. Leeb1, G. R. M¨uller-Putz1, A. Schl¨ogl, and G.Pfurtscheller | journal & 2008.}
• Evaluation Criteria for BCI Research]
  _{MIT PRess |2007}
• Sleep Physionet dataset- Alexander Grahmfort

• Corpus

• Physionet.org/
• A list of all public EEG-datasets.
• Big list(pdf) of Publicly Available EEG Datasets

General AI papers

• Meta-Transfer Learning for Few-Shot Learning]
  _{Qianru Sun1,3∗ Yaoyao Liu2∗ Tat-Seng Chua1 Bernt Schiele3 CVPR & Year.}
• Model-agnostic meta-learning for fast adaptation of deep networks
  _{Chelsea Finn,Pieter Abbeel,Sergey Levine - ICML & 08/2017.}
• A concise review of recent few-shot meta-learning methods]
  _{Xiaoxu Li a,1, Zhuo Sun b,1, Jing-Hao Xue b,⇑, Zhanyu Ma c ,Neurocomputing & 28/10/2020.}
• Meta-learning approaches for learning-to-learn in deep learning: A survey]
  _{Yingjie Tian a,b,c,⇑, Xiaoxi Zhao a,b,c, Wei Huang | Neurocomputing & 21/04/2022.}
• Comparing Transfer and Meta Learning Approaches on a Unified Few-Shot Classification Benchmark
• A Unified Few-Shot Classification Benchmark to Compare Transfer and Meta Learning Approaches
• Blog - Meta Leaning is all you need
• Blog - Summary of Few shot Learning and Meta learning concepts

• Metaleanring in other similar domain

• Few Is Enough: Task-Augmented Active Meta-Learning for Brain Cell Classification
• ENHANCING GENERALIZATION OF FIRST-ORDER META-LEARNING - Reptile updated version -2019
• On First-Order Meta-Learning Algorithms

• MetaL Applicaiton

  • Meta-ED: Cross-lingual Event Detection Using Meta-learning for Indian Languages

• TL in other similar domain

  • Boosting for Transfer Learning

Thesis

• FEW-SHOT CLASSIFICATION OF EEG WITH QUASI-INDUCTIVE TRANSFER LEARNING | Bachelor thesis

Awesome-Githubs

• MetaEEG
• berdakh
• Transfer Learning Algorithms for EEG-based BCI
• Motor Imagery Tutorial
• EEG Project Example: EEG Face Detection
• Machine learning on electrophysiology MEG/EEG signals Course at MAIN educational 2022 - Agrahmfort and Hubert Banville
• EEGModels project: A Collection of Convolutional Neural Network (CNN)
• SSVEP pipeline UCSD Challange
• Key packages for Signal Processing
• ML-DL for EEG- INTERFACES*-BRAIN lab (to be explored)
• pbashivan/EEGLearn
• kylemath/EEGEdu

• Researcher Githubs

• Vinay Jayaram meta (TL)
• Felix Heilmeyer - Freiburg University Medical Center
• Data analyst, Brain researcher, Pain psychology neuroscientist
• Aung Aung Phyo Wai -- researcher NTU Nanyang Technological University Singapor

• Few-Shot Github

• Few-Shot Neuroimaging(Few-Shot leanring for Brain activation map) \

Conferences

• DeepBCI

Books

• NEUROSCIENCE EXPLORING THE BRAIN - has good brief EEG details
• Brain Computer Interface handbook- Fabien Lotte, Chang Nam, Anton Nijholt --2018

BCi Starter Learning Courses or Tutotials

• TNT Academy
• brain Encoding Decoding Conferences
• Deep Learning for EEGs and BCI : (How I Managed to get 94% score model
• Deep Learning for EEGs and BCI : (How I Managed to get 94% score model.)
• Colab - Tutorial on EEG processing to classification
• Colab Notebook - Braindecode_Tutorial.ipynb

Kaggle Compitition on EEG

• Grasp-and-Lift EEG Detection Identify hand motions from EEG recordings

Awesome YouTube Playlists

• Design BCI project - Hardware, Software , Analyssi and ML
• Getting Started with BCI Projects Series
• EEG ML/DL Youtube full coding only Playlist(has a lot of code)
• PyTorch for Brain-Computer Interfaces
• Intro to Brain-Computer Interfaces (Winter 2023)
• Tutorial on ML and EEG/MEG by Alexander grahmfort

Resources-for-newbie_ResearchStudents

• Research and Writing Tips from SJN University
• Notes on writing Fredo Durand, MIT CSAIL
• Write Good Papers Frédo Durand
• Resources for Students & Scholars by Frédo Durand; MIT

Awesome-tools

• Visulaisation Data-to-viz
• EEG2image Synchrosqueezing in Python

---

DL4EEG - newbie resource

• Review of the BCI competition IV
• Deep learning with convolutional neural networks for EEG decoding and visualization

DL4BCI Researcher \ google Scholars

• Robin Tibor Schirrmeister- PhD Candidate, Neuromedical AI Lab and Machine Learning Lab, University Freiburg

Lab

• Brain & behaviour Lab

---

so the next section is my brainstorming place for futur eof BCI - This is in idation phase and you find it intersting or have something to share then I would love to hear you- please mail me at [email protected]

Ai4BCI - An Open AI tool box to bring BCI out of the lab

Goal - bridge the gap and develope resources which are more accessible. We are planning to start this initiative with the aim of making BCI an everyday technology, just like AI has made computer vision, NLP, and speech processing accessible to everyone. Every hand now has access to these technologies, but BCI, despite being an earlier development, remains confined to the lab. Why is that?

but Why?

• if not now then when? ;)
• lot of companies are really focusing on finding people now who are in this field but there is need of acommunity that foster them really well
• there is no set major for this field
• BCI, a really cool area but really underdeveloped despite of all the hype
• having had a bit of experience lookoing at these larger companies and seeing the direction they're heading at the end of the day what the intention for all of this would be to develope the kind of systems or develop more approachable ways to understand BCI as a whole if it's going to become a technology that is widely used in the future. it seems that it is going toi be one of the technologies that will be likely used in our generations- it's important that we grasp basic understanding of how to build smaller project , tutorial, curriculam, communities around technology that will be impactful.

---

---